Harvard Journal of Law & Technology
Volume 16, Number 2 Spring 2003
ADVERSE POSSESSION FOR INTELLECTUAL PROPERTY:
ADAPTING AN ANCIENT CONCEPT TO RESOLVE CONFLICTS
BETWEEN ANTITRUST AND INTELLECTUAL PROPERTY
LAWS IN THE INFORMATION AGE
Constance E. Bagley† and Gavin Clarkson‡
TABLE OF CONTENTS
I. INTRODUCTION ................................................................................... 328
II. IP AND ANTITRUST BASICS................................................................ 330
A. Patents, Copyrights, and Trade Secrets ........................................ 330
B. Antitrust Statutes ........................................................................... 332
C. Monopolization ............................................................................. 333
D. Duty to Deal.................................................................................. 334
1. Essential Facilities Doctrine....................................................... 334
2. Changes in Preexisting Business Practices................................. 336
E. Monopoly Leveraging.................................................................... 336
F. Tying.............................................................................................. 338
G. Exclusive Dealing Arrangements.................................................. 341
III. ORIGINS OF THE INTELLECTUAL PROPERTY RIGHT
REFUSAL TO DEAL QUANDARY .......................................................... 341
A. Kodak I .......................................................................................... 342
B. Kodak II......................................................................................... 344
C. Xerox ............................................................................................. 346
D. Microsoft....................................................................................... 350
E. Other Relevant Cases .................................................................... 356
1. Other Patent Cases ..................................................................... 357
2. Other Copyright Cases ............................................................... 359
IV. MAKING SENSE OF KODAK, XEROX, AND MICROSOFT ....................... 363
A. Our Proposal................................................................................. 365
1. Relevant Justifications for Adverse Possession ......................... 366
2. Firms Without Market Power..................................................... 373
3. Firms with Market Power........................................................... 375
† Associate Professor of Business Administration at the Harvard Business School. JD,
magna cum laude, Harvard Law School. Partner, Bingham McCutchen, 1984–1990.
‡ Olin Research Fellow in Law, Economics, and Business at the Harvard Business
School. JD, cum laude, Harvard Law School. The authors wish to thank John Allison, Kristen Carpenter, Joe Kalt, Mark Lemley, Jeff MacKie-Mason, Lynda Oswald, Joe Singer,
Robert Williams, and the peer reviewers in the 2003 Huber Hurst Seminar on Legal Research at the University of Florida for helpful comments and suggestions. Any errors or
omissions are solely the fault of the authors.
328
Harvard Journal of Law & Technology
[Vol. 16
B. Application to Unilateral Refusal to Deal..................................... 377
1. Large Fixed Assets (the Kodak and Xerox scenarios) ................ 377
2. Creation of a New Product Market: Segway.............................. 378
3. Operating Systems and APIs (Microsoft)................................... 379
4. European Application................................................................. 381
C. Application to Unilateral but Conditional Refusal to
Deal ............................................................................................ 383
1. Large Fixed Assets..................................................................... 384
2. Segway ....................................................................................... 384
3. Operating Systems ..................................................................... 384
V. POSSIBLE CRITIQUES ......................................................................... 385
A. Real Property Critique — Why Not Easements for
Intellectual Property?................................................................. 385
B. Intellectual Property and Antitrust Critiques................................ 389
VI. CONCLUSION .................................................................................... 391
I. INTRODUCTION
Is a unilateral1 refusal to license or sell intellectual property immune from a claim of monopolization or attempted monopolization
under the Sherman Act? Does the answer change if the product in
question is a new or an existing product? Does it matter whether the
intellectual property rights are protected by copyright rather than patent? Although in 1992 the Supreme Court addressed these issues
somewhat obliquely in Kodak v. Image Technical Services (“Kodak
I”),2 these questions assumed new urgency in 2000 after the Federal
Circuit created a split in the circuits by ruling in CSU v. Xerox that
patent holders “may enforce the statutory right to exclude others from
making, using, or selling the claimed invention free from liability un-
1. By unilateral, we mean non-collusive. See BLACK’S LAW DICTIONARY 1532 (7th ed.
1999); see also HERBERT HOVENKAMP, MARK D. JANIS, & MARK A. LEMLEY, IP AND
ANTITRUST: AN ANALYSIS OF ANTITRUST PRINCIPLES APPLIED TO INTELLECTUAL
PROPERTY LAW (2002), §13.4 (distinguishing unilateral refusals from conditional and/or
concerted refusals) [hereinafter IP AND ANTITRUST].
2. Eastman Kodak Co. v. Image Technical Servs., Inc., 504 U.S. 451 (1992) [hereinafter
Kodak I]. Within the Court’s discussion of tying, footnote 29 of that opinion states that
“power gained through some natural and legal advantage such as a patent, copyright, or
business acumen can give rise to liability if ‘a seller exploits his dominant position in one
market to expand his empire into the next.’” Id. at 480, n.29 (citations omitted). At the time,
however, intellectual property rights were not an issue before the court. See Jonathan Gleklen, “Antitrust Liability for Unilateral Refusal to License Intellectual Property: Xerox and its
Critics,” 6, unpublished manuscript available at http://www.ftc.gov/opp/intellect/020501
gleklen.pdf (stating that “[F]ootnote 29 can be characterized as dicta because the Kodak
case did not involve the rights of intellectual property owners — the only evidence before
the Court was that none of Kodak’s parts were patented.”).
No. 2]
Adverse Possession for Intellectual Property
329
der the antitrust laws.”3 The Federal Circuit expressly declined4 to
follow a Ninth Circuit ruling that held antitrust liability could be imposed for almost identical conduct, depending on the motivation of
the patent holder.5 The District of Columbia Circuit’s statement in
United States v. Microsoft (“Microsoft 2001”)6 that Microsoft’s assertion that its copyrights on its software programs entitled it to impose
restrictions on its original equipment manufacturer (“OEM”) licensees
was “no more correct than the proposition that use of one’s personal
property, such as a baseball bat, cannot give rise to tort liability”7 only
added fuel to the firestorm of controversy.
The depth of this quandary became even more evident when, presented with a seemingly obvious split in the circuits, the Solicitor
General suggested that the U.S. Supreme Court deny certiorari in the
Xerox case to “allow these difficult issues to percolate further in the
courts of appeals.”8 Indeed, the Antitrust Division of the Justice Department and the Federal Trade Commission held an entire day of
hearings on exactly these questions as part of their series of hearings
on competition and intellectual property policy.9 The hearings produced a substantial amount of thought and discussion, but no clear
solution.
This Article focuses on two related questions. First, under what
circumstances must the holder of a patent or a copyright or the owner
of a trade secret allow others to use that intellectual property? Second,
under what circumstances can the holder of an intellectual property
right use that right to make it difficult for another party to succeed in a
related market?
Given that determining antitrust liability for refusing to deal is already “one of the most unsettled and vexatious [questions] in the anti3. CSU, LLC v. Xerox Corp. (In re Independent Service Organizations Antitrust Litigation), 203 F.3d 1322, 1327 (Fed. Cir. 2000), cert. denied, 531 U.S. 1143 (2001) [hereinafter
Xerox].
4. See id.
5. See Image Technical Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195 (9th Cir.
1997), cert. denied, 523 U.S. 1094 (1998), [hereinafter Kodak II].
6. United States v. Microsoft Corp., 253 F.3d 34, 62 (D.C. Cir. 2001) [hereinafter Microsoft 2001].
7. Id. at 63.
8. Brief for the United States as Amicus Curiae at 16, Xerox, 531 U.S. 1143 (2001) (No.
00-62), available at http://www.usdoj.gov/osg/briefs/2000/2pet/6invit/2000-0062.pet.ami.
inv.html.
9. Department of Justice Antitrust Division & Federal Trade Commission, Hearing on
Competition and Intellectual Property Law and Policy in the Knowledge Based Economy,
"Strategic Use of Licensing: Is There Cause For Concern about Unilateral Refusals to
Deal?" May 1, 2002, available at http://www.ftc.gov/opp/intellect/020501xscript.pdf.; see
also R. Hewitt Pate, Acting Assistant Attorney General, Antitrust Division, U.S. Department of Justice, Address before the American Intellectual Property Law Association, (Jan.
24, 2003), available at http://www.usdoj.gov/atr/public/speeches/200701.htm (last visited
April 2, 2003) ("[T]he joint DOJ/FTC IP hearings . . . examined a number of issues [including unilateral] refusals to license intellectual property").
330
Harvard Journal of Law & Technology
[Vol. 16
trust field,”10 the additional requirement of determining the scope of
the intellectual property grant and the concomitant right to exclude
makes this problem that much more complex.11 In order to arrive at a
potential solution, it is necessary to retrace briefly the jurisprudential
path to see how this issue arose. Part II of this Article discusses some
of the intellectual property and antitrust principles that should guide
the development of any potential solution.12 Part III then reviews the
origin and propagation of the problem of intellectual property rights
(“IPR”) holders unilaterally refusing to deal. In Part IV we present our
proposed solution, which calls for the application of the real property
concept of adverse possession in the intellectual property arena, along
with a slight extension of the Essential Facilities Doctrine for industries that exhibit network effects. Our proposed solution is designed to
allow these cases to be resolved via summary judgment, and we demonstrate this potential judicial efficiency by applying our framework
to various potential situations involving refusals to deal. In Part V we
attempt to answer some potential critiques of our solution, and in Part
VI we conclude with suggestions for further exploration.
II. IP AND ANTITRUST BASICS
A. Patents, Copyrights, and Trade Secrets
Intellectual property is any product or result of a mental process
that is given legal protection against unauthorized use.13 Intellectual
property law identifies four categories of innovation — patents, copyrights, trade secrets, and trademarks — and provides separate protection for each.14
Although the present controversy focuses on patents and copyrights because they are the most likely forms of IP to be at issue in a
refusal to license, refusal to disclose trade secrets has also been a
problem in the antitrust context. As part of the consent decree settling
the suit brought by the United States and a number of individual
10. Byars v. Bluff City News Co., 609 F.2d 843, 846 (6th Cir. 1979).
11. Determinations of patent scope have huge economic implications. Josh Lerner has
empirically demonstrated that patent scope significantly affects firm value. See Joshua
Lerner, The Importance of Patent Scope: An Empirical Analysis, 25 RAND J. ECON. 319
(1994); see also Robert P. Merges & Richard R. Nelson, On the Complex Economics of
Patent Scope, 90 COLUM. L. REV. 839 (1990).
12. Certain aspects of the analysis and arguments presented in Part II were first introduced in Constance E. Bagley, “United States v. Microsoft: Application of the Antitrust
Laws to the New Economy,” Harvard Business School, N9-802-090 (Sept. 20, 2002).
13. See generally CONSTANCE E. BAGLEY, MANAGERS AND THE LEGAL ENVIRONMENT:
STRATEGIES FOR THE 21ST CENTURY (4th ed. 2002), particularly Chapter 11.
14. Certain countries, such as the United States, have created additional areas of specialized intellectual property protection such as photolithography masks and plant variety patents.
No. 2]
Adverse Possession for Intellectual Property
331
states,15 Microsoft Corporation is obligated to reveal certain trade secrets to competitors, for instance, providing early disclosure of its
application programming interfaces (“APIs”).16
A copyright includes the government-granted right to prevent
others from copying the original expression embodied in a work.17
The owner has exclusive rights to distribute, display, and perform the
work, and to create derivative works.18 It is the expression that is protected, not the ideas underlying the expression.19 The exclusivity of
copyright protection also has limits; certain “fair use” defenses are
permitted, such as copying portions of the copyrighted work for
“criticism, comment, news reporting, teaching (including multiple
copies for classroom use), scholarship, or research.”20
A patent is the government-granted right that protects any “useful
process, machine, manufacture, or composition of matter, or any new
and useful improvement thereof”21 that is both novel22 and nonobvious.23 After a period of time the patent expires and the invention
is dedicated to the public. The patent holder need not personally make
use of the invention during the life of the patent to enjoy protection.
Patents do not entitle the patent holder to make or sell the patented
invention. Instead, what is granted is a right to exclude others from
making, using, or selling the same invention. The exclusivity of the
patent grant is not unlimited, however. The doctrine of patent misuse
keeps the patent owner from overreaching and attempting to do more
than is legitimately authorized under the patent grant. Most of the patent misuse cases24 take the form of fraudulent procurement of a patent,25 bad faith enforcement of a patent,26 or some other form of
15. United States v. Microsoft Corp., 231 F. Supp. 2d 144 (D.D.C. 2002) [herinafter
Microsoft 2002].
16. Consent Decree, § III.D, New York v. Microsoft Corp., 224 F. Supp. 2d 76, 266
(D.D.C. 2002).
17. See 17 U.S.C. § 102 (2000) (protecting literary works, musical works, sound recordings, audiovisual works, sculptural works, pictorial works, computer programs, or any
other original work of authorship fixed in a tangible medium of expression).
18. See 17 U.S.C. § 106 (2000).
19. See 17 U.S.C. § 102(b) (2000).
20. See 17 U.S.C. § 107 (2000). The statute specifies four factors for evaluating fair use:
(1) the purpose and character of the use, including whether such use
is of a commercial nature or is for non-profit educational purposes;
(2) the nature of the copyrighted work;
(3) the amount and substantiality of the portion used in relation to the
copyrighted work as a whole; and
(4) the effect of the use upon the potential market for or value of the
copyrighted work. Id.
21. See 35 U.S.C. § 101 (2000).
22. See 35 U.S.C. § 102 (2000).
23. See 35 U.S.C. § 103 (2000).
24. See DONALD S. CHISUM, ET AL., PRINCIPLES OF PATENT LAW 937–76 (discussing
patent misuse).
25. See Walker Process Equip., Inc. v. Food Mach. & Chem. Corp., 382 U.S. 172 (1965).
332
Harvard Journal of Law & Technology
[Vol. 16
antitrust violation involving the patent.27 Unlike copyright, there is no
“fair use” defense to patent infringement. “[E]xperimental use” of a
patented invention is permitted but is “very narrow and strictly limited” to instances where the actions performed are “for amusement, to
satisfy idle curiosity, or for strictly philosophical inquiry.”28
Unlike copyrights and patents, which are based on federal law,
trade secrets are protected by state laws, most of which are based on
the Uniform Trade Secrets Act.29 A trade secret is any process, innovation, or information that gives its owner a competitive advantage by
reason of its being kept secret.30 The owner of a trade secret has the
right to prohibit anyone from engaging in the unauthorized use or disclosure of the trade secret for as long as the owner takes reasonable
steps to keep it confidential.31
Given that the intellectual property laws grant exclusive rights to
the holders of patents, copyrights, and trade secrets, it is not surprising
that a regime with a predisposed hostility to monopolies, such as the
antitrust system, would find aspects of the intellectual property regime
problematic.
B. Antitrust Statutes
Section 1 of the Sherman Act provides that “[e]very contract,
combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several states, or with foreign
nations, is declared to be illegal.”32 Although the language in the statute is unequivocal, it has been interpreted to bar only “unreasonable”
restraints on trade. As then-Judge Breyer explained: “Virtually every
contract to buy ‘forecloses’ or ‘excludes’ alternative sellers from some
portion of the market, namely the portion consisting of what was
bought.”33
Section 2 of the Sherman Act makes it a felony for a firm to “monopolize, or attempt to monopolize, or combine or conspire with any
other person or persons, to monopolize any part of the trade or commerce among the several States.”34 The offense of monopolization has
26. See Handgards, Inc. v. Ethicon, Inc., 601 F.2d 986 (9th Cir. 1979), cert. denied, 444
U.S. 1025 (1980).
27. We will examine whether a refusal to license a patent constitutes an antitrust violation in Parts III and IV infra.
28. Madey v. Duke Univ., 307 F.3d 1351, 1362 (Fed. Cir. 2002) (quoting Embrex Inc. v.
Service Eng’g Corp., 216 F.3d 1343, 1349 (Fed. Cir. 2000)).
29. See, e.g., CAL. CIV. CODE § 3426 (2003). Note that this and many other state statutes
are based on the Unif. Trade Secrets Act (1985).
30. See, e.g., CAL. CIV. CODE § 3426.1(D) (2003).
31. See, e.g., CAL. CIV. CODE § 3426.2 (2003).
32. 15 U.S.C. § 1 (2000).
33. Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d 227, 236 (1st Cir. 1983) (emphasis in original).
34. 15 U.S.C. § 2 (2000).
No. 2]
Adverse Possession for Intellectual Property
333
two elements: “(1) the possession of monopoly power in the relevant
market and (2) the willful acquisition or maintenance of that power as
distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.”35
C. Monopolization
The U.S. Supreme Court defines monopoly power as “the power
to control prices or exclude competition” in the relevant market.36
“Because the ability of consumers to turn to other suppliers restrains a
firm from raising prices above the competitive level,”37 the relevant
market includes all products “reasonably interchangeable by consumers for the same purposes.”38 Definition of the relevant product market
is critical. The broader the definition, the less dominant one firm’s
position will be.39
Having a monopoly does not by itself violate Section 2. A Section
2 violation occurs only when a firm acquires or maintains, or attempts
to acquire or maintain, a monopoly by engaging in anticompetitive
conduct.40 As Judge Learned Hand explained, “The successful competitor, having been urged to compete, must not be turned upon when
he wins.”41 Anticompetitive conduct includes refusal to deal and other
exclusionary practices. In many cases, it is difficult to determine
whether any particular act is exclusionary or merely a form of vigorous competition.
To be condemned as exclusionary, a monopolist’s act must have
an “anticompetitive effect” that harms the competitive process and
thereby consumers. Harm to competitors, even if motivated by malice,
is not enough.42 The plaintiff bears the burden of demonstrating that
the monopolist’s conduct has the requisite anticompetitive effect. If
the plaintiff successfully demonstrates anticompetitive effect, then
the monopolist may proffer a “procompetitive justification” for its conduct. If the monopolist asserts a
procompetitive justification — a nonpretextual claim
35. United States v. Grinnell Corp., 384 U.S. 563, 570–71 (1966).
36. United States v. E. I. du Pont de Nemours & Co., 351 U.S. 377, 391 (1956).
37. Rothery Storage & Van Co. v. Atlas Van Lines, Inc., 792 F.2d 210, 218 (D.C. Cir.
1986).
38. E.I. du Pont de Nemours, 351 U.S. at 395.
39. See Microsoft 2001, 253 F.3d 52. (Microsoft argued, unsuccessfully, that the relevant
market was not just Intel-compatible PC operating systems but also middleware (such as
Java or Internet browsers), hand-held computers, and the Apple Computer operating system.).
40. See Grinnell, 384 U.S. at 571.
41. United States v. Aluminum Co. of Am., 148 F.2d 416, 430 (2d Cir. 1945).
42. See Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 225
(1993).
334
Harvard Journal of Law & Technology
[Vol. 16
that its conduct is indeed a form of competition on
the merits because it involves, for example, greater
efficiency or enhanced consumer appeal — then the
burden shifts back to the plaintiff to rebut that
claim.43
If the monopolist’s procompetitive justification stands unrebutted,
the plaintiff must demonstrate that the anticompetitive harm of the
conduct outweighs the procompetitive benefit. Finally, in considering
whether the monopolist’s conduct on balance harms competition and
is therefore condemned as exclusionary under Section 2, courts focus
on the effect of that conduct, not upon the intent behind it.44
D. Duty to Deal
In general, even a monopolist does not have a duty to deal with
others. Unilateral refusals to deal are most likely to be problematic
when they are used to “punish” others for doing business with a competitor.45 The Patent Act expressly provides that refusing to license
one’s patented property or imposing conditions on licenses of such
property is not itself unlawful in most circumstances.46 There are,
however, at least two instances where there is a duty to deal.
1. Essential Facilities Doctrine
A monopolist has a duty to deal with its rivals when it controls an
essential facility — a resource necessary to its rivals’ survival that
they cannot feasibly duplicate. To prevail on an essential facilities
claim, a plaintiff must establish that (1) access to the facility is essential to other firms’ ability to compete; (2) the facility is controlled by a
monopolist; (3) the plaintiff cannot practically or reasonably duplicate
the facility; (4) the plaintiff has been denied the use of the facility; and
(5) it is feasible to provide the facility to the plaintiff.47
The earliest application of the Essential Facilities Doctrine involved a group of railroads that used the same bridge over the Mississippi River to transport freight and passengers to St. Louis. Several of
43. Microsoft 2001, 253 F.3d at 59 (citations omitted).
44. But see discussion of Kodak II, Part III.B infra.
45. See, e.g., Consent Decree, § III.D, New York v. Microsoft Corp., 224 F. Supp. 2d 76,
266 (D.D.C. 2002) (prohibiting Microsoft from charging OEMs higher prices for Microsoft’s Windows operating system if those OEMs installed Netscape Navigator and mandating that Microsoft provide uniform license terms to certain OEMs).
46. 35 U.S.C. § 271(d) (2002) (“No patent owner otherwise entitled to relief . . . shall be
denied relief or deemed guilty of misuse or illegal extension of the patent right by reason of
his having . . . refused to license or use any rights to the patent”).
47. See MCI Communications Corp. v. Am. Tel. & Tel. Co., 708 F.2d 1081, 1132–33
(7th Cir. 1983), cert. denied, 440 U.S. 971 (1983).
No. 2]
Adverse Possession for Intellectual Property
335
the railroads bought the company that controlled the bridge and terminal at St. Louis, then refused to permit the other railroads to use the
bridge even though it was the only bridge in the area. The U.S. Supreme Court held that the bridge constituted an essential facility that
had to be made available, on reasonable terms, to all of the railroads.48
This approach was followed in the case against American Telephone and Telegraph (“AT&T”), which led to the breakup of the Bell
Telephone System.49 AT&T refused to give access to local telephone
lines and related equipment to rivals hoping to enter the long-distance
market, making it very difficult for competitors to offer long-distance
service in competition with AT&T’s high-price service.
In practice, courts are reluctant to characterize products or facilities as “essential,” particularly when there is no horizontal element
involving competitors.50 The Essential Facilities Doctrine does not
require a firm to share resources that are merely useful. A company
need not share technology that would allow competitors to compete
more effectively or resources that competitors could duplicate on their
own. For example, the Seventh Circuit held that the New York Times
News Service (the only service containing New York Times articles
and crossword puzzles) was not an essential facility because there
were at least three major competing supplemental news services.51
Similarly, the U.S. Court of Appeals for the Tenth Circuit held
that membership in Visa USA’s umbrella organization for issuers of
Visa credit cards was not an essential facility to which Sears, the issuer of Discover Cards, had to be given access.52 As a result, Visa
could refuse to admit as members any financial institution that issued
Discover or American Express cards. Unlike the pre-existing railroad
facilities at issue in Terminal Railroad, Visa USA had spent years
creating a brand and operating systems that Sears and American Express “not only had done nothing to create, but had chosen to compete
against.”53 The bylaw restricting membership (1) prevented Sears and
American Express from free-riding in a market in which there was no
evidence of increased prices or reduced output; (2) did not bar Sears
48. United States v. Terminal R.R. Ass’n of St. Louis, 224 U.S. 383 (1912) (Terminal
Railroad).
49. United States v. Am. Tel. & Tel. Co., 552 F. Supp. 131 (D.D.C. 1982), aff’d sub nom.
Maryland v. United States, 460 U.S. 1001 (1983). As with Microsoft, the ultimate result in
the AT&T case was a consent decree rather than a decision on the merits. See id. at 135;
Microsoft 2002, 231 F. Supp. 2d 144, 149 (D.D.C. 2002).
50. Thus, one could conceivably limit the holding in Terminal Railroad to cases where at
least two direct competitors agree to exclude others from the use of a facility. See 2 JULIAN
O. VON KALINOWSKI ET AL., ANTITRUST LAWS AND TRADE REGULATION § 25.04 n.113 (2d
ed. 2002) and accompanying text.
51. Paddock Publ’ns, Inc. v. Chicago Tribune Co., 103 F.3d 42, 44 (7th Cir. 1996), cert.
denied, 520 U.S. 1265 (1997) (stating that “a newspaper deprived of access to the New York
Times crossword puzzles can find others, even if the Times has the best known one”).
52. SCFC ILC, Inc. v. Visa USA, Inc., 36 F.3d 958 (10th Cir. 1994).
53. Id. at 970.
336
Harvard Journal of Law & Technology
[Vol. 16
from access to the general-purpose credit card market; and (3) did not
alter the character of the general-purpose credit market or change any
present pattern of distribution.54 Ultimately, the restrictions were
struck down in a later antitrust action brought by the Department of
Justice. However, this action was not based on an application of the
Essential Facilities Doctrine but instead on a finding that the exclusionary rules restricted competition and harmed consumers by denying them innovative and varied products.55
2. Changes in Preexisting Business Practices
In certain instances courts have found violations of Section 2
when a firm with market power acted to change a pre-existing distribution pattern. In Aspen Skiing Co. v. Aspen Highlands Skiing
Corp.,56 the U.S. Supreme Court closely scrutinized an instance where
a “monopolist elected to make an important change in a pattern of
distribution that had originated in a competitive market and had persisted for several years.”57 The Court found this sudden refusal to deal
exclusionary because the monopolist was not motivated by efficiency
concerns and was willing to sacrifice short-run benefits and consumer
goodwill in exchange for a perceived long-run impact on its “smaller
rival.”58 Other courts have also found that to impose antitrust liability,
a change in longstanding business practices must have occurred.59
E. Monopoly Leveraging
Ordinarily, Section 2 liability is restricted to monopolistic behavior within the specific market in which the firm has monopoly power.
Through leveraging, however, a firm with monopoly power in one
market can gain an advantage in a separate market. A firm clearly
violates Section 2 when it uses that advantage to attain monopoly
power in the second market. It is less clear whether a firm can legally
use monopoly power in one market to gain a competitive advantage,
short of actual monopolization, in another market.
In Berkey Photo, Inc. v. Eastman Kodak Co. (“Berkey Photo”), a
film seller and processor challenged Kodak’s refusal to provide the
specifications for a new pocket-sized camera system requiring a new
type of film and processing.60 Berkey Photo accused Kodak of at54. Id. at 971–72.
55. United States v. Visa U.S.A., Inc., 163 F. Supp. 2d 322 (S.D.N.Y. 2001).
56. 472 U.S. 585 (1985).
57. Id. at 603.
58. Id. at 610.
59. See, e.g., SMS Sys. Maint. Servs., Inc. v. Digital Equip. Corp., 11 F. Supp. 2d 166,
168 (D. Mass. 1998) (finding no liability without change in conduct).
60. 603 F.2d 263 (2d Cir. 1979), cert. denied, 444 U.S. 1093 (1980).
No. 2]
Adverse Possession for Intellectual Property
337
tempting to use its monopoly power in the film market to gain leverage in the photo-finishing equipment and services markets in violation
of Section 2 of the Sherman Act. The Second Circuit rejected
Berkey’s claim and concluded that Kodak’s invention resulted from
its superior business skill, product, and foresight. Thus, Kodak’s refusal to disclose the product innovation prior to the introduction of the
new camera did not constitute willful maintenance of monopoly
power in violation of the Sherman Act.61
The Berkey Photo court did, however, suggest that using monopoly power in one market to obtain a competitive advantage in another
might violate Section 2.62 Recent U.S. Supreme Court decisions tend
to undercut this suggestion, but not decisively. More specifically, in
Spectrum Sports, Inc. v. McQuillan, the Supreme Court held that the
plaintiff in a Section 2 case involving unilateral conduct must prove a
dangerous probability that the defendant’s conduct will create or
maintain monopoly power in a market.63 That holding suggests leveraging that confers only a competitive advantage in a market cannot
violate Section 2.
The U.S. Court of Appeals for the Ninth Circuit adopted this approach when it held that two airlines that had developed the two largest proprietary computerized airline reservation systems did not
violate Section 2 merely because the systems gave them a competitive
advantage in the air-transportation market by listing their flights
first.64 The court held that unless the monopolist uses its power in the
first market to acquire and maintain a monopoly in the second market,
or attempts to do so, there is no Section 2 violation. The plaintiffs
conceded that the two airlines did not have a monopoly in the airtransportation market and that there was no dangerous probability that
either defendant would acquire such a monopoly. Therefore, the court
rejected their Section 2 claims.
On the other hand, language in Kodak I suggests that leveraging
monopoly power in one market to create a competitive advantage in
another market can be a Section 2 violation.65 Therefore, it is not yet
clear what position the Supreme Court would take on the legality of
monopoly leveraging that confers only a competitive advantage.
61. Id.
62. Id.
63. See 506 U.S. 447 (1993).
64. See Alaska Airlines, Inc. v. United Airlines, Inc., 948 F.2d 536 (9th Cir. 1991), cert.
denied, 503 U.S. 977 (1992).
65. See 504 U.S. at 483; see also discussion in Part III infra.
338
Harvard Journal of Law & Technology
[Vol. 16
F. Tying
Tying arrangements can be challenged under Section 1 of the
Sherman Act66 or, in cases involving commodities, under Section 3 of
the Clayton Act.67 In a “tying arrangement,” the seller will sell product A (the “tying,” or desired, product) to the customer only if the
customer agrees to purchase product B (the “tied” product). A tying
arrangement forces a buyer to purchase a product or service that it
would not buy based solely on the product’s or service’s own merits.
There are four elements to a per se tying violation: (1) the tying
and tied goods must be two separate products; (2) the defendant must
have market power in the tying product market; (3) the defendant
must afford consumers no choice but to purchase the tied product
from it; and (4) the tying arrangement must foreclose a not insubstantial volume of commerce in the tied product.68
The primary concern behind the prohibition on tying “is that tying
prevents goods from competing directly for consumer choice on their
merits, i.e., being selected as a result of buyers’ independent judgment.”69 Direct competition on the merits of the tied product is foreclosed when the tying product either is sold only in a bundle with the
tied product or, though offered separately, is sold at a bundled price,
so that the buyer pays substantially less for the bundled product than
he or she would pay for the tying product and tied products if purchased separately (an “economic tie”). In either case, the consumer
becomes less willing to buy a competitor’s version of the tied product
even if the competitor’s version is superior in terms of quality and
price.
As the Supreme Court explained in Jefferson Parish, the “answer
to the question whether one or two products are involved [does not
turn] on the functional relation between them.”70 The focus is instead
on whether there is sufficient consumer demand for the tied product,
separate from the tying product, to create a distinct product market in
which it is efficient for competitors to offer only the tied product.71
The Court concluded that a hospital that conditioned surgical care at
its facility on the purchase of anesthesiological services from an affiliated medical group sold two separate products — anesthesia, the
tied service, and surgical care, the tying service — even though patients rarely had surgery without an anesthetic or an anesthetic with-
66. 15 U.S.C. § 1 (2000).
67. 15 U.S.C. § 14 (2000).
68. See, e.g., Indep. Ink, Inc. v. Trident, Inc., 210 F. Supp. 2d 1155, 1162 (C.D. Cal.
2002).
69. Microsoft 2001, 253 F.3d 52, 87 (D.C. Cir. 2001) (internal quotation marks omitted).
70. Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 19 (1984).
71. See id. at 21–22.
No. 2]
Adverse Possession for Intellectual Property
339
out surgery.72 In other words, the mere fact that two items are complements, that one “is useless without the other,”73 does not make
them a single product for purposes of tying law.
The Court examined both the direct and the indirect evidence of
consumer demand for the tied product separate from demand for the
tying product. For direct evidence, the Court looked at whether, given
a choice, consumers purchased the tied product from the maker of the
tying product or from other firms.74 The Court found that patients and
surgeons often requested specific anesthesiologists not associated with
the hospital where the surgery was to be performed.75 For indirect
evidence, the Court looked at the behavior of firms without market
power in the tying product and found that only 27 percent of anesthesiologists had financial relationships with hospitals.76
As the Court of Appeals for the District of Columbia Circuit explained in its Microsoft 2001 opinion, there is always, in the abstract,
direct separate demand for products: If choice is available at zero cost,
consumers will prefer choice to no choice.77 But when the benefits of
bundling outweigh the benefits of choice, most consumers will not
choose to make independent purchases.
The consumer demand test is a “rough proxy for whether a tying
arrangement may, on balance, be welfare-enhancing, and unsuited to
per se condemnation.”78 Thus, unlike per se violations, a tying arrangement may be upheld if there is a business justification for it.79
For example, courts have upheld the right of a franchisor to require
franchisees to purchase certain supplies as part of the franchise
agreement: Courts have required Baskin-Robbins franchisees to buy
ice cream only through authorized Baskin-Robbins distribution channels80 and required Domino’s Pizza franchisees to purchase Domino’s-approved supplies and ingredients.81 According to the Court of
Appeals for the Third Circuit, both
72. Id. at 22.
73. Id. at 19 n.30.
74. Id. at 22–23.
75. Id. at 22.
76. Id. at 22 n.36.
77. Microsoft 2001, 253 F.3d at 87.
78. Id.
79. See, e.g., Mozart Co. v. Mercedes-Benz of N. Am., Inc., 833 F.2d 1342 (9th Cir.
1987), cert. denied, 488 U.S. 870 (1988) (Mercedes-Benz’s policy of requiring its dealers to
sell only factory-made parts was found to be justified by the assurance it provided to Mercedes that service on its automobiles, important in preserving a high-quality image, would
not be performed with substandard parts.); see also discussion of Data Gen. Corp. v.
Grumman Sys. Support infra Part III.E.
80. See Krehl v. Baskin-Robbins Ice Cream Co., 664 F.2d 1348 (9th Cir. 1982) (BaskinRobbins trademark lacked sufficient independent existence apart from the branded ice
cream products to justify a finding of an unlawful tying arrangement.).
81. See Queen City Pizza v. Domino’s Pizza, 124 F.3d 430 (3d Cir. 1997) (Plaintiffs’ acceptance of a franchise package with full knowledge that the franchise contract included
purchase requirements and contractual restrictions could not give rise to antitrust liability
340
Harvard Journal of Law & Technology
[Vol. 16
courts and legal commentators have long recognized
that franchise tying contracts are an essential and
important aspect of the franchise form of business
organization because they reduce agency costs and
prevent franchisees from free riding . . . . We do not
believe the antitrust laws were designed to erect a serious barrier to this form of business organization.82
As long as potential franchisees are aware of the contractual restrictions requiring the purchase of supplies only from approved sources,
their “remedy, if any, is in contract, not under the antitrust laws.”83
In United States v. Microsoft Corp. (“Microsoft 1998”),84 the
Court of Appeals for the District of Columbia Circuit considered
whether bundling Internet Explorer with Windows 95 constituted an
“integrated product” for purposes of a 1995 consent decree.85 In the
decree, Microsoft had agreed not to condition the sale of one product
on the sale of another but had retained the right to develop “integrated
products.”86 The court held that two products are “integrated” if providing them as a bundle offers the consumer functionalities not available if the products were purchased separately.87 Because the
Windows 95 and Internet Explorer bundle did offer distinct functionalities, the court concluded that they were integrated.88 The court distinguished “bolting,” which is technologically linking two products to
make functions unavailable to purchasers of the separate products,
from “integration,” which creates a whole greater than the sum of its
parts.89 Although the court indicated in Microsoft 1998 that it believed
the separate product analysis would have been the same under the
antitrust laws,90 it backed away from this statement in Microsoft 2001
and limited its prior holding to the language of the 1995 consent decree itself.91
when a competitive market for franchises existed in which numerous alternative franchise
opportunities were available. “If the contractual restrictions in . . . the general franchise
agreement were viewed as overly burdensome or risky at the time they were proposed,
plaintiffs could have purchased a different form of restaurant, or made some alternative
investment.”). Id. at 441.
82. Id. at 440–41.
83. Id. at 441.
84. United States v. Microsoft Corp., 147 F. 3d 935 (D.C. Cir. 1998) [hereinafter Microsoft 1998].
85. United States v. Microsoft Corp., 1995 U.S. Dist. LEXIS 20533, 8 (D.D.C. 1995).
86. Id.
87. Microsoft 1998, 147 F.3d 935.
88. Id.
89. Id. at 949.
90. Id. at 950.
91. Microsoft 2001, 253 F.3d 34, 92 (D.C. Cir. 2001); see also discussion infra Part
IV.B.3.
No. 2]
Adverse Possession for Intellectual Property
341
G. Exclusive Dealing Arrangements
An exclusive dealing arrangement does not violate the antitrust
laws unless its probable effect is to foreclose “competition . . . in a
substantial share of the line of commerce affected.”92 Accordingly,
unless a contract forecloses competition in a substantial part of the
market for a substantial period of time, an exclusive dealing arrangement does not violate Section 1. For example, the Seventh Circuit
upheld the New York Times News Service’s practice of selling the
reprint rights to its crossword puzzles and other features to only one
newspaper in a particular geographic market in any given twelvemonth period.93 Here, the exclusive contract was of short duration and
left other bidders free to obtain the rights by bidding the next year.
In upholding the exclusive contract, Judge Easterbrook wrote:
“Competition-for-the-contract is a form of competition that antitrust
laws protect rather than proscribe, and it is common.”94 Exclusive
stories and features help newspapers differentiate themselves and thus
better compete with each other, especially “if smaller newspapers are
willing to bid with cash rather than legal talent.”95 Firms such as the
New York Times News Service would, he argued, have less incentive
to find and explicate the news if the law prevented them from deciding how to best market their intellectual property for maximum
profit.96 In addition, consumers may benefit from having different
articles and features available in different newspapers.
III. ORIGINS OF THE INTELLECTUAL PROPERTY RIGHT
REFUSAL TO DEAL QUANDARY
If an original manufacturer makes spare parts for its equipment
generally available, enterprising equipment service firms (“Independent Service Organizations” or “ISOs”) may create an ancillary market
for servicing the equipment. Should an original manufacturer wish to
control the service market for its equipment, it might cut off the supply of parts to the ISOs. These were the circumstances surrounding
Kodak I,97 a case that spawned several generations of ISO litigation.
92. Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320, 327 (1961).
93. Paddock Publ’ns, Inc. v. Chicago Tribune Co., 103 F.3d 42 (7th Cir. 1996), cert. denied 520 U.S. 1265 (1997).
94. Id. at 45.
95. Id.
96. Id.
97. Kodak I, 504 U.S. 451 (1992).
342
Harvard Journal of Law & Technology
[Vol. 16
A. Kodak I
Although best known for its photographic business, Kodak also
manufactures and sells photocopiers and micrographic equipment.98
Kodak also provides service and spare parts to its customers.99 Kodak
directly produces some of these parts, and it uses OEMs to produce
the rest.100 During the early 1980s, ISOs began repairing and servicing
Kodak equipment at prices substantially lower than those charged by
Kodak.101 Some ISO customers purchased their own parts and hired
the ISOs to provide the service, while others used the ISOs for both
parts and service.102 The ISOs either obtained parts from Kodak or the
OEMs, or they scavenged used parts and refurbished them.103
In 1985, Kodak implemented a policy of selling replacement copier parts only to customers who purchased service from Kodak or who
repaired their own machines.104 In 1986, Kodak implemented a similar policy for its micrographic equipment products.105 In both instances, Kodak also mandated that the OEMs follow a similar policy
of not selling to ISOs.106 When their sources of spare parts began to
evaporate, many ISOs were forced out of business; those that managed to survive lost substantial revenue.107
A number of ISOs filed an antitrust suit in 1987 alleging that Kodak was in violation of Section 1 of the Sherman Act by tying the sale
of copier service to the sale of parts.108 The ISOs also alleged that
Kodak violated Section 2 by unlawfully monopolizing and attempting
to monopolize the sale of service for Kodak equipment.109
The District Court for the Northern District of California granted
Kodak’s motion for summary judgment.110 In a split decision, the
Court of Appeals for the Ninth Circuit reversed.111 On certiorari,112
the Supreme Court held that products manufactured by a single
firm — in this case, Kodak copier parts — could constitute the relevant market.113 This is most likely to be the case when the costs of
98. Kodak’s micrographic equipment products include microfilm scanners, printers,
viewers, and retrieval systems. See id. at 457 n.1.
99. See id. at 455.
100. See id. at 457.
101. See id.
102. See id. at 458.
103. See id. at 458 n.2.
104. See id. at 458.
105. See Kodak II, 125 F.3d 1195, 1201 (9th Cir. 1997).
106. See Kodak I, 504 U.S. at 458.
107. See id.
108. Id. at 459.
109. Id.
110. Id.
111. Id. at 460.
112. Kodak I, Inc., 501 U.S. 1216 (1991) (mem.).
113. Kodak I, 504 U.S. at 481–82.
No. 2]
Adverse Possession for Intellectual Property
343
switching to another supplier’s product are very high and the lack of
availability of the product from other sources is not known in advance.114 Kodak did not have monopoly power in the copier market.
Nonetheless, because the cost of replacing the copier was high, both
absolutely and relative to the costs of parts and service, consumers
were unlikely to discard their Kodak copier when offered replacement
parts only as part of a service contract. 115
Because of these high switching costs, the Court held that the
relevant product market for the purposes of summary judgment was
the market for Kodak copier replacement parts.116 As sole provider of
Kodak parts, Kodak clearly had monopoly power in that market. The
Supreme Court concluded that Kodak’s policy of selling replacement
parts only to buyers of Kodak equipment who use Kodak repair service could be an illegal tie-in sale, as well as monopolization or an
attempt to monopolize the service and parts markets in violation of
Section 2, unless there was a valid business justification for the policy. 117 As guidance for remand, the Court stated:
In the end, of course, Kodak’s arguments may prove
to be correct. It may be that its parts, service, and
equipment are components of one unified market, or
that the equipment market does discipline the aftermarkets so that all three are priced competitively
overall, or that any anticompetitive effects of Kodak’s behavior are outweighed by its competitive effects.118
Although the issue of whether Kodak’s conduct was within the
scope of the patent grant was not before the Court,119 the Court included in its opinion a footnote that became quite important in later
cases. Footnote 29 states, in part:
[We have] held many times that power gained
through some natural and legal advantage such as a
patent, copyright, or business acumen can give rise
to liability if a seller exploits his dominant position
in one market to expand his empire into the next.120
114. Id. at 473–74.
115. Id. at 476–77.
116. Id. at 473.
117. Id. at 483–84.
118. Id. at 486.
119. See Gleklen, supra note 2, at 6 n.21.
120. Kodak I, 504 U.S. at 479 n.29 (internal citations omitted).
344
Harvard Journal of Law & Technology
[Vol. 16
Like Aspen Skiing,121 Kodak I focused on the anticompetitive effects of a change in longstanding business practices.122 Kodak I has
been cited for the principle that a unilateral change in policy can constitute anticompetitive behavior.123
B. Kodak II
The case was remanded to the district court. At trial, Kodak asserted for the first time that a valid business justification for its anticompetitive conduct was its desire to protect copyrighted and patented
parts.124 While the Supreme Court had noted in Kodak I that a refusal
to deal is justified if there are “legitimate competitive reasons for the
refusal,”125 the district court found that the proffered business justification was pretextual.126 Kodak’s parts manager “testified that patents
‘did not cross [his] mind’ at the time Kodak began the parts policy.
Further, no distinction was made by Kodak between ‘proprietary’
parts covered by tooling or engineering clauses and patented or copyrighted products.”127 In addition, the district court noted that, although
Kodak equipment “requires thousands of parts, . . . only 65 were patented. [Thus] this case concern[ed] a blanket refusal that included
protected and unprotected products.”128
Before closing arguments, the ISOs dropped their Section 1 tying
claims, and thus only the Section 2 attempted monopolization and
monopolization claims were submitted to the jury.129 The jury
121. See Part II.D.2, supra.
122. See generally HERBERT HOVENKAMP, FEDERAL ANTITRUST POLICY §7.5 at 265–66
(1994) (recommending that Aspen Skiing and Kodak I be applied only to the cancellation of
participation in a pre-existing joint venture, not to create a new obligation to deal where no
arrangement had existed before); Daniel M. Wall, Aftermarket Monopoly Five Years After
Kodak, 11 ANTITRUST 32, 37 (Summer 1997) (“As substantive law, [Kodak I’s] lasting
significance may be to carry forward antitrust’s peculiar animosity, usually associated with
Aspen Skiing, towards changes in business policies.”); Michael S. Jacobs, Market Power
Through Imperfect Information: The Staggering Implications of Eastman Kodak Co. v.
Image Technical Services and a Modest Proposal for Limiting Them, 52 MD. L. REV. 336,
360 n.129 (1993) (Aspen Skiing establishes that a monopolist may not change, without valid
business justification, an established course of cooperative dealing with a rival).
123. See, e.g., Queen City Pizza v. Domino’s Pizza, 124 F.3d 430, 440 (3d Cir. 1997).
(“[Kodak I] arose out of concerns about unilateral changes in Kodak’s parts and repairs
policies. . . . Because this change in policy was not forseen at the time of sale, buyers had no
ability to calculate these higher costs at the time of purchase and incorporate them into their
purchase decision.”); see also discussion supra Part II.D.2.
124. Kodak II, 125 F.3d at 1212. Kodak held 220 U.S. patents “covering 65 parts for its
[equipment], and all Kodak diagnostic software and service software [was] copyrighted.” Id.
at 1214.
125. Kodak I, 504 U.S. at 483 n.32.
126. Kodak II, 125 F.3d at 1219.
127. Id.
128. Id.
129. Id. at 1201.
No. 2]
Adverse Possession for Intellectual Property
345
awarded treble damages of $71.8 million, and the court imposed a tenyear injunction requiring Kodak to sell “all parts” to ISOs.130
Kodak’s appeal returned the case to the Court of Appeals for the
Ninth Circuit. While the court of appeals acknowledged that a desire
to protect intellectual property rights may create a legitimate business
justification for a refusal to deal, in Kodak II it found Kodak’s defense
to be pretextual.131 The court held that “a monopolist’s ‘desire to exclude others from its [intellectual property] is a presumptively valid
business justification for any immediate harm to consumers’”132 resulting from that monopolist’s unilateral refusal to license a patent or
copyright or to sell its patented or copyrighted work. That presumption may be rebutted by evidence that the intellectual property protection was acquired unlawfully or that its use as a business justification
was merely pretext.133
The court of appeals recognized the tension between antitrust
law’s concerns about monopoly as a threat to competition and the
public policy underlying the copyright and patent laws granting limited monopolies as an incentive to innovate.134 The court acknowledged that patent and copyright holders may refuse to sell or license
their protected work, even if the result is to give them a monopoly on
the patented or copyrighted products.135 The court noted that it could
“find no reported case in which a court has imposed antitrust liability
for a unilateral refusal to sell or license a patent or copyright.”136
However, the court of appeals cited Kodak I for the proposition
that a monopolist who acquires a dominant position in one market
through patents and copyrights may violate Section 2 if it exploits that
dominant position to extend the lawful monopoly into a separate market.137 According to the court of appeals, Kodak did not have a valid
business justification for its policies and therefore violated Section 2
when it used its monopoly power in the Kodak copier parts market to
obtain a monopoly in the Kodak copier service market.138
130. Id. at 1200.
131. Id. at 1219. The trial court had instructed the jury that Kodak’s intellectual property
rights provided no defense if its actions otherwise constituted monopoly leveraging. Id. at
1214. Although the Court of Appeals for the Ninth Circuit held that this jury instruction was
erroneous, it found this error to be harmless because Kodak’s defense was pretextual. Id. at
1218–20.
132. Id. at 1218 (quoting Data Gen. Corp. v. Grumman Sys. Support, 36 F.3d 1147, 1197
(1st Cir. 1994)).
133. Id. at 1218.
134. Id. at 1215.
135. Id. at 1216.
136. Id.
137. Id. (citing Kodak I, 504 U.S. at 480 n.29).
138. Id. at 1220.
346
Harvard Journal of Law & Technology
[Vol. 16
C. Xerox
During the same period as the Kodak litigation, Xerox also manufactured, sold, and provided service for photocopiers. In 1984, Xerox
“established a policy of not selling parts unique to its series 10 copiers
to [ISOs] unless they were also end-users of the copiers.”139 Xerox
expanded the policy in 1987 “to include all new products as well as
existing series 9 copiers.”140
In contrast to the policy implemented by Kodak, however, Xerox
end-users “remained free to supply the parts they purchased to any
service provider, including ISOs,” as long as the parts were ultimately
installed in end-users’ copiers.141 Additionally, all ISOs could still
purchase parts at the same prices with the same quantity discounts as
any other end-user142 as long as the parts were installed in their own
machines.143 Although Xerox limited the quantity of parts to any
given end-user to the “quantity reasonably necessary to repair the
[end-user’s] equipment, [end-users] were free to resell the parts they
purchased to ISOs.”144
At issue, however, was a change in Xerox policy. In 1989, Xerox
tightened its policy and “implemented an ‘on-site end-user verification’ procedure” to confirm that the parts ordered for end-users were
actually installed in end-users’ copiers.145 A group of ISOs filed a
class action lawsuit against Xerox, which Xerox settled in 1994.146
Under the terms of the settlement, Xerox “agreed to suspend its restrictive parts policy for six and one-half years and to license its diagnostic software for four and one-half years.”147 CSU opted out of that
settlement and filed its own antitrust lawsuit alleging that Xerox’s
refusal to sell its patented parts or to license its copyrighted manuals
and software violated the Sherman Act.148 Unlike the plaintiff in Kodak I, however, CSU did not make any allegations of tying.
Xerox responded by “arguing that its prior unilateral refusal to
sell patented parts and to license patented and copyrighted diagnostic
software (and its post-settlement pricing of such items) were a lawful
unilateral exercise of its intellectual property rights and hence could
139. Xerox, 203 F.3d 1322, 1324 (Fed. Cir. 2000), cert. denied, 531 U.S. 1143 (2001).
140. Id.
141. Gleklen, supra note 2, at 2.
142. See id. The sole exception involved the U.S. Navy, which had a special contract
whereby Xerox provided service to shipboard copiers when the vessels were in port. Navy
technicians trained by Xerox serviced the copiers when at sea. Id. at 2 n.4.
143. See Xerox, 203 F.3d at 1324.
144. Gleklen, supra note 2, at 2.
145. See Xerox, 203 F.3d at 1324.
146. Id.
147. Id.
148. See id. at 1325.
No. 2]
Adverse Possession for Intellectual Property
347
not violate § 2 [of the Sherman Act].”149 In an attempt to circumvent
the issue of intent highlighted by the Ninth Circuit in Kodak II,
“Xerox did not claim that its refusal to deal was in fact motivated by
its intellectual property rights, but [instead] asserted that the existence
of such IP rights immunized its conduct.”150 Xerox also argued that
“CSU could not show it was actually injured by Xerox’s refusal to sell
[unpatented] parts.”151 Thus any injury was caused solely by “Xerox’s
lawful refusal to sell or license patented parts and copyrighted software.”152
Xerox also counterclaimed for patent and copyright infringement
“based on CSU’s use of unlicensed diagnostic software and infringing
parts purchased from third party parts vendors.”153 CSU “defended the
infringement with a patent and copyright misuse defense based on the
same conduct that underlay its antitrust claims — Xerox’s prior refusal to sell or license and its current pricing of parts and software.”154
As in Kodak I, the district court granted summary judgment and
dismissed CSU’s antitrust claims, “holding that if a patent or copyright is lawfully acquired, [the] unilateral refusal to sell or license . . .
is not unlawful exclusionary conduct under the antitrust laws, even if
[such a refusal] impacts competition in more than one market.”155 The
appeal was taken by the Federal Circuit because, at least since Nobelpharma,156 antitrust issues involving the scope of the patent grant
were viewed as being within the Federal Circuit’s exclusive jurisdiction.157
149. Gleklen, supra note 2, at 3.
150. Id.
151. Id.
152. Xerox, 203 F.3d at 1324.
153. Gleklen, supra note 2, at 3.
154. Id.
155. Xerox, 203 F.3d at 1324.
156. Nobelpharma AB v. Implant Innovations, Inc., 141 F.3d 1059 (Fed. Cir. 1998); see
also Midwest Indus., Inc. v. Karavan Trailers, Inc., 175 F.3d 1356, 1360 (Fed. Cir. 1999)
(en banc in relevant part) (“Pro-Mold and Nobelpharma make clear that our responsibility as
the tribunal having sole appellate responsibility for the development of patent law requires
that we do more than simply apply our law to questions of substantive patent law. In order
to fulfill our obligation of promoting uniformity in the field of patent law, it is equally important to apply our construction of patent law to the questions whether and to what extent
patent law preempts or conflicts with other causes of action.”).
157. See Xerox, 203 F.3d at 1325. However, after Holmes Group v. Vornado Air Circulation Sys., 535 U.S. 826 (2002), it is unclear whether such a case would automatically continue to fall within the Federal Circuit’s exclusive jurisdiction. In Holmes, the Supreme
Court held that the well-pleaded complaint rule requires that for a case to arise under patent
law, the plaintiff's complaint must establish either that federal patent law itself created the
cause of action or that “the plaintiff's right to relief necessarily depends on resolution of a
substantial question of federal patent law.” Id at 1893, quoting Christianson v. Colt Indus.
Operating Corp., 486 U.S. 800, 809 (1988). If cases involving antitrust immunity for activities within the scope of the patent grant are to remain the exclusive jurisdiction of Federal
Circuit, it will likely have to be under this second rationale.
348
Harvard Journal of Law & Technology
[Vol. 16
In its brief to the Federal Circuit, Xerox argued that its “patents
claim inventions that [made] servicing copiers and printers easier and
more effective.”158 For example, “as noted in Xerox’s brief in the
Federal Circuit, Xerox’s ‘797 pressure roll patent notes that invention
provides the advantage of ‘less down-time required for replacing pressure rolls,’ i.e., the need for less frequent copier and printer service.”159 Additionally, “Xerox’s ‘256 fuser roll patent states that the
object of this invention is to provide a new fuser roll with a longer
lasting coating, a goal that is desirable because it ‘is expensive to . . .
install fuser rolls.’”160 Xerox’s ‘006 patent for a “document handler
belt . . . disclosed a belt with ‘good document handling performance
without requiring the intermittent or periodic adjustment required’ by
servicemen for proper operation of the prior art belt.” 161
The Federal Circuit held that Xerox could refuse to sell its patented replacement parts to independent service providers such as
CSU.162 In doing so, the Federal Circuit expressly declined to follow
the Ninth Circuit’s position in Kodak II that determination of the
defendant’s subjective motivation is warranted to determine whether
the defendant’s actions really were motivated by a desire to exploit its
intellectual property.
The Federal Circuit acknowledged that intellectual property rights
do not confer a privilege to violate antitrust laws, and the D.C. Circuit
cited Xerox with approval for this proposition in Microsoft 2001.163
But unlike the D.C. Circuit, the Federal Circuit went on to state that as
a general rule, the antitrust laws do not prevent the owner of intellectual property rights from excluding others from use of its patented
property.164 A patent owner who sues or countersues to exclude others
from making, using, or selling the claimed invention is exempt from
antitrust laws unless (1) the “patent was obtained through knowing
and willful fraud” on the Patent and Trademark Office, or (2) the patent infringement suit or counter-claim is a “mere sham” to disguise an
attempt “to interfere directly with the business relationships of a competitor.”165 Neither of these two exceptions was at issue in the case.166
158. Gleklen, supra note 2, at 13, n.51 and accompanying text.
159. Id. at 13, n.51.
160. Id. Note also that the non-stick coating used for fuser rolls also has application in
non-stick cookware. “The invention also has utility in the field of coating metal substrates,
for example in the production of cooking utensils and other surfaces used in the culinary
arts . . . to produce surfaces which provide release.” U.S. Patent No. 4,196,256 (issued Apr.
1, 1980).
161. Gleklen, supra note 2, at 13, n.51.
162. Xerox, 203 F.3d at 1328.
163. Microsoft 2001, 253 F.3d 34, 63 (D.C. Cir. 2001).
164. Xerox, 203 F.3d 1325 (citing Intergraph Corp. v. Intel Corp., 195 F.3d 1346, 1362
(Fed. Cir. 1999)).
165. Id. at 1326.
166. Id. Of course, these exceptions only apply to activity that is within the scope of the
patent. Id. at 1327–28.
No. 2]
Adverse Possession for Intellectual Property
349
The Federal Circuit rejected CSU’s reliance on the Supreme
Court’s suggestion in Kodak I that a seller cannot exploit its dominant
position in one market to expand into another market.167 The Federal
Circuit characterized Kodak I as a tying case168 and suggested that
when “[p]roperly viewed within the framework of a tying case, [footnote 29 from Kodak I] can be interpreted as restating the undisputed
premise that the patent holder cannot use his statutory right to refuse
to sell patented parts to gain a monopoly in a market beyond the scope
of the patent.”169 Unlike the plaintiff in Kodak I, CSU had made no
claims that Xerox had tied the sale of its patented parts to its unpatented products or services.170
The Federal Circuit also declined to inquire into the subjective
motivation behind Xerox’s refusal to sell or license its patented
works.171 Instead, the court reiterated that unless the patent infringement claims are objectively baseless, “an antitrust defendant’s subjective motivation is immaterial.”172 As long as the antitrust defendant
did not engage in illegal tying, Patent and Trademark Office fraud, or
sham litigation, it could enforce its statutory intellectual property
rights without violating the antitrust laws.173
With respect to Xerox’s refusal to license its copyrighted manuals
and software, the court cited Supreme Court precedent emphasizing
that property rights “granted by copyright law cannot be used with
impunity to extend power into the marketplace beyond what Congress
intended.”174 As the Ninth Circuit had done in Kodak II, the Federal
Circuit held that “an author’s desire to exclude others from use of its
copyrighted work is a presumptively valid business justification for
any immediate harm to consumers” resulting from the exclusionary
conduct.175 But unlike the Ninth Circuit, the Federal Circuit declined
to examine Xerox’s subjective motivation for relying on the copyright
laws to refuse to sell its copyrighted materials.176 As it had held with
respect to patented property, the Federal Circuit held that the antitrust
plaintiff can rebut the presumption of validity only by proving that the
antitrust defendant had obtained the copyrights by unlawful means or
167. See id. at 1327.
168. See id. (“Notably, [Kodak I] was a tying case when it came before the Supreme
Court, and no patents had been asserted in defense of the antitrust claims against Kodak.”).
169. Id.
170. See id. (“[T]here are no claims in this case of illegally tying the sale of Xerox's patented parts to unpatented products. Therefore, the issue was not resolved by the Kodak
language cited by CSU.”).
171. See id.
172. Id. (quoting Nobelpharma AB v. Implant Innovations, Inc., 141 F.3d 1059, 1072
(Fed. Cir. 1998) (internal quotation marks omitted)).
173. Id. at 1327.
174. Id. at 1328 (citing United States v. Loew’s, Inc., 371 U.S. 38, 47–48 (1962)).
175. Id. at 1329 (quoting Data Gen. Corp. v. Grumman Sys. Support Corp., 36 F.3d
1147, 1187 (1st Cir. 1994) (internal quotation marks omitted)).
176. Id.
350
Harvard Journal of Law & Technology
[Vol. 16
used them to gain monopoly power beyond the statutory copyright
granted by Congress.177
D. Microsoft178
While this debate over unilateral refusals to deal was raging, Microsoft Corporation179 was embroiled in defending itself against
sweeping allegations by the U.S. Justice Department and nineteen
states that Microsoft had maintained a monopoly in the market for
Intel-compatible personal computer (“PC”) operating systems through
several means, including illegally tying Internet Explorer to Windows
95 and Windows 98 in violation of Section 1.180 The plaintiffs also
alleged that Microsoft entered into unlawful exclusive dealing arrangements in violation of Section 1 and unlawfully attempted monopolization of the internet browser market in violation of Section
2.181 The states brought similar claims under analogous state laws.182
The case arose, in large part, out of “Microsoft’s varied efforts to unseat Netscape Navigator as the preeminent internet browser.”183 Less
than a year and a half after the Federal Circuit decision in Xerox, the
D.C. Circuit rendered its unanimous decision in the Microsoft antitrust case.184
The district court had determined that Microsoft had (1) maintained a monopoly in the market for Intel-compatible PC operating
systems in violation of Section 2 of the Sherman Act; (2) attempted to
gain a monopoly in the market for internet browsers in violation of
177. Id.
178. Portions of this discussion are drawn from Bagley, supra note 12.
179. Microsoft is the largest computer software company in the world and creator of the
Windows operating system used on more than 95 percent of all Intel-compatible personal
computers. See United States v. Microsoft Corp., 87 F. Supp. 2d 30, 36 (D.D.C. 2000)
[hereinafter Microsoft 2000].
180. Microsoft 2001, 253 F.3d 34, 47 (D.C. Cir. 2001). Other allegations included (1) entering into restrictive licensing agreements with OEMs, such as Hewlett Packard, that prohibited OEMs from (a) replacing desktop icons, folders, and Start folder entries; (b)
modifying the Windows start-up sequence; (c) launching Windows with any interface other
than the Windows desktop; or (d) adding any icons, in size or shape, different from those
provided by Microsoft; (2) including Internet Access Providers, such as America Online, in
the Windows Online Service sign-up folder in exchange for their using Internet Explorer as
the default browser; (3) giving preferential treatment to Internet Content Providers and
Independent Software Vendors in exchange for their using Internet Explorer as the default
browser; (4) entering into an exclusive dealing agreement with Apple Computer; and (5)
undermining Sun Microsystem’s Java middleware by (a) creating an incompatible Java
Virtual Machine; (b) inducing Independent Software Vendors to write only for the Microsoft version; (c) deceiving programmers; and (d) coercing Intel to stop supporting Java. See
generally id. at 34. This case was discussed in Part II.F, supra.
181. Id. at 47.
182. Id.
183. Id. at 46.
184. Microsoft 2001 was decided on June 28, 2001. Xerox was decided on February 17,
2000.
No. 2]
Adverse Possession for Intellectual Property
351
Section 2; and (3) illegally tied two purportedly separate products,
Windows and Internet Explorer, in violation of Section 1.185 The district court also found that the same facts that established liability under the Sherman Act mandated findings of liability under analogous
state law antitrust provisions.186 To remedy the Sherman Act violations, the district court issued a final judgment requiring Microsoft to
submit a proposed plan of divestiture, with the company to be split
into an operating systems business and an applications business.187
On appeal, the D.C. Circuit acknowledged that it was deciding
this case
against a backdrop of significant debate amongst
academics and practitioners over the extent to which
“old economy” § 2 monopolization doctrines should
apply to firms competing in dynamic technological
markets characterized by network effects. In markets
characterized by network effects, one product or
standard tends towards dominance, because “the utility that a user derives from consumption of the good
increases with the number of other agents consuming
the good.” For example, “an individual consumer’s
demand to use (and hence her benefit from) the telephone network . . . increases with the number of
other users on the network whom she can call or
from whom she can receive calls.” Once a product or
standard achieves wide acceptance, it becomes more
or less entrenched. Competition in such industries is
“for the field” rather than “within the field.”
In technologically dynamic markets, however, such
entrenchment may be temporary, because innovation
may alter the field altogether. Rapid technological
change leads to markets in which “firms compete
through innovation for temporary market dominance,
from which they may be displaced by the next wave
of product advancements.”188
185. Microsoft 2000, 87 F. Supp. 2d 30, 35 (D.D.C. 2000).
186. Id.
187. Id. at 59.
188. Microsoft 2001, 253 F.3d 34, 49 (D.C. Cir. 2001) (citations omitted); see also Bruce
Abramson, From Investor Fantasy to Regulatory Nightmare: Bad Network Economics and
the Internet’s Inevitable Monopolists, 16 HARV. J.L. & TECH. 159, 173 (2002) (“The first
and best-known defining characteristic of a network industry is that the value of a network
grows with its size. This phenomenon is known as ‘positive feedback’ (or alternatively,
increasing returns to scale).” The phenomenon of positive feedback “follows standard microeconomic principles describing consumer choice. The rational choice for a consumer . . .
352
Harvard Journal of Law & Technology
[Vol. 16
In particular, the court noted that “there is no consensus among
commentators on the question of whether, and to what extent, current
monopolization doctrine should be amended to account for competition in technologically dynamic markets characterized by network
effects.”189
The court also pointed out the practical difficulty of crafting
remedies in cases involving violations of the antitrust laws in technologically dynamic markets.190 Although the court deemed it “noteworthy” that a case of this magnitude and complexity could proceed from
the filing of complaints to an appellate decision in a “mere three
years,” the court noted that more than six years had passed since Microsoft engaged in the first alleged anticompetitive conduct, a period
of time that “seems like an eternity in the computer industry.”191 As a
result:
Conduct remedies may be unavailing in such cases,
because innovation to a large degree has already rendered the anticompetitive conduct obsolete (although
by no means harmless). And broader structural
remedies present their own set of problems, including how a court goes about restoring competition to a
dramatically changed, and constantly changing, marketplace.192
Even though the forward-looking remedies available in government enforcement actions involving technologically dynamic markets
“appear[ed] limited,” the court indicated that enforcement actions
would still play an “important role” in “defining the contours of the
antitrust laws so that law-abiding firms will have a clear sense of what
is permissible and what is not.” 193 The court went on to state that “the
threat of private damage actions will remain to deter those firms inclined to test the limits of the law.”194
The D.C. Circuit (1) upheld the district court’s determination that
Microsoft’s exclusive dealing contracts with Internet Access Providers did not violate Section 1;195 (2) affirmed in part and reversed in
part the district court’s judgment that Microsoft violated Section 2 by
is to select the option that will connect her to the most valuable of the competing networks. .
. . [Because of the higher utility associated with joining a larger network,] the consumer’s
rational decision to join it will make the largest network larger — and thus correspondingly
more valuable to the next consumer who considers joining.”).
189. Id. at 50.
190. Id.
191. See id. at 48–49.
192. Id. at 49.
193. Id.
194. Id.
195. Id. at 70 (“Plaintiffs did not cross-appeal this holding.”).
No. 2]
Adverse Possession for Intellectual Property
353
employing anticompetitive means to maintain a monopoly in the operating system market;196 (3) reversed the district court’s determination that Microsoft violated Section 2 by illegally attempting to
monopolize the internet browser market;197 (4) remanded the finding
that Microsoft violated Section 1 by unlawfully tying its browser to its
operating system;198 and (5) vacated the remedial order calling for
divestiture because (a) the district court’s final judgment rested on a
number of liability determinations that did not survive appellate review; (b) the district court failed to hold an evidentiary hearing to address remedy-specific factual disputes; and (c) the trial judge had
“engaged in impermissible ex parte contacts by holding secret interviews with members of the media and made numerous offensive
comments about Microsoft officials in public statements outside of the
courtroom, giving rise to an appearance of partiality.”199
Although the D.C. Circuit’s opinion covered a wide range of alleged transgressions, and at times exonerated Microsoft, we focus on
the aspects of the opinion addressing the OEM licensing restrictions
and the alleged tying of Windows and Internet Explorer.200
The D.C. Circuit upheld the district court’s finding of monopoly
power in a relevant market, defined as Intel-compatible PC operating
systems, rejecting Microsoft’s attempt to define the market more
broadly.201 Microsoft had a greater than 95 percent share of that market and its market position was protected by a substantial entry barrier, namely the “applications barrier to entry.”202 That barrier
stemmed from the fact that “(1) most consumers of software prefer
operating systems for which a large number of applications have already been written; and (2) most developers prefer to write for operating systems that already have a substantial consumer base.”203
The appeals court also held that “with the exception of the one restriction prohibiting automatically launched alternative interfaces, all
196. Id. at 46.
197. Id.
198. Id.
199. Id.
200. The district court held that Microsoft’s exclusive dealing contracts with Internet Access Providers (such as AOL) did not violate Section 1, because they did not completely
exclude Netscape from reaching any potential user by some means of distribution, such as
downloading Netscape’s browser software from the Internet or distributing free software
disks. The plaintiffs did not appeal this ruling. The government also had alleged that Microsoft’s pricing of Internet Explorer was predatory. By pricing Internet Explorer below cost
(indeed, by even paying people to take it), Microsoft was able to preserve its stream of
monopoly profits on Windows and thereby more than recoup its investment in below-cost
pricing on Internet Explorer. The district court did not assign liability for predatory pricing.
Because the plaintiff did not assert this theory on appeal, the court of appeals did not consider it. Id. at 68.
201. Id. at 52.
202. Id. at 55.
203. Id.
354
Harvard Journal of Law & Technology
[Vol. 16
the OEM license restrictions at issue represent uses of Microsoft’s
market power to protect its monopoly, unredeemed by any legitimate
justification” in violation of Section 2.204 Microsoft had attempted to
defend its OEM license restrictions as lawful conditions on the rights
it gave others to use its intellectual property. The court rejected Microsoft’s “bold and incorrect” position that if intellectual property
rights have been lawfully acquired then their subsequent exercise cannot give rise to antitrust liability, stating that this assertion was “no
more correct than the proposition that use of one’s personal property,
such as a baseball bat, cannot give rise to tort liability.”205 As noted
earlier, the D.C. Circuit cited with approval language from the Federal
Circuit’s opinion in Xerox: “Intellectual property rights do not confer
a privilege to violate the antitrust laws.”206
The D.C. Circuit did, however, accept Microsoft’s argument that
even a copyright holder with monopoly power may legally limit a
licensee’s ability to engage in “significant and deleterious alterations
of a copyrighted work.”207 Even so, the court concluded that the only
license restriction necessary to prevent a substantial alteration of Microsoft’s copyrighted work was a restriction on OEMs’ ability to
automatically launch a substitute user interface upon completion of
the boot process.208 The court agreed “that a shell that automatically
prevents the Windows desktop from ever being seen by the user is a
drastic alteration of Microsoft’s copyrighted work, and outweighs the
marginal anticompetitive effect of prohibiting the OEMs from substituting a different interface automatically upon completion of the initial boot process.”209 None of the other elements of the ruling
regarding product design specifically implicated intellectual property
rights.210
204. Id. at 64.
205. Id. at 62.
206. Id. at 63 (quoting Xerox, 203 F.3d 1322, 1325 (Fed. Cir. 2000)). Note, however, that
the D.C. Circuit declined to include the subsequent statement from the Xerox opinion (“But
it is also correct that the antitrust laws do not negate the patentee's right to exclude others
from patent property”). Perhaps this is because the intellectual property at issue was based
on copyright rather than a patent. Whether such a divergent outcome should depend on
whether the asset in question is a copyright or a patent is discussed in Part IV infra.
207. Microsoft 2001, 253 F.3d 34, 63 (D.C. Cir. 2001) (quoting Gilliam v. ABC, 538
F.2d 14, 21 (2d Cir. 1976)).
208. Id. at 63. The court found that the primary justifications put forward by Microsoft
“border[ed] upon the frivolous.” Id.
209. Id.
210. The appeals court upheld the district court’s conclusion that Microsoft’s exclusion
of Internet Explorer from the Add/Remove Programs utility and its commingling of browser
and operating system code constituted unjustified exclusionary conduct. The appeals court
reversed, however, the finding that using Windows to override the user’s choice of a default
browser in certain circumstances constituted unjustified exclusionary conduct, after concluding that the plaintiffs had failed to rebut Microsoft’s proffered “valid technical reasons”
for the override. Id. at 67.
No. 2]
Adverse Possession for Intellectual Property
355
Turning to the claims of attempted monopolization, the appeals
court reversed the finding that Microsoft illegally attempted to monopolize the Internet browser market, because the plaintiffs had failed
to show that the browser market could be monopolized, “i.e., that a
hypothetical monopolist in that market could enjoy market power.”211
In particular, the plaintiffs failed “(1) to define the relevant market,
and (2) to demonstrate that substantial barriers to entry protect that
market.”212
In perhaps the most groundbreaking part of the opinion, the D.C.
Circuit deviated from the Supreme Court’s holding that tying is per se
illegal if there are separate products and the defendant has market
power in the tying product.213 Instead, the D.C. Circuit ruled that “the
rule of reason, rather than per se analysis, should govern the legality
of tying arrangements involving platform software products.”214 Notwithstanding its ruling in Microsoft 1998 that, at least for purposes of
construing the 1995 consent decree, Windows and Internet Explorer
were not separate products,215 the Microsoft 2001 court determined
The appeals court also upheld the finding that Microsoft’s exclusive dealing contracts
with the Internet Access Providers, such as AOL, were unjustified exclusionary devices in
violation of Section 2. Id. at 71. The appeals court reversed, however, the district court’s
finding that the dealings with the Internet Content Providers violated Section 2, because
there was insufficient evidence to support a finding that Microsoft’s promotional restrictions
actually had a substantial, deleterious impact on Navigator’s usage share. The appeals court
upheld, however, the finding that the exclusive dealing arrangements with the independent
service vendors and Apple Computer constituted unjustified exclusionary conduct in violation of Section 2. Id. at 74.
As for Java, the appeals court held that the development of the incompatible Java Virtual
Machine did not violate Section 2, reasoning that “a monopolist does not violate the antitrust laws simply by developing a product that is incompatible with those of its rivals. In
order to violate the antitrust laws, the incompatible product must have an anticompetitive
effect that outweighs any procompetitive justification for the design.” Id. at 75. Because the
Microsoft Java Virtual Machine allowed applications to run more swiftly on Windows,
Microsoft did not violate the antitrust laws when it developed and promoted its own Java
Virtual Machine. On the other hand, the court upheld the finding that the provisions requiring use of Microsoft’s Java Virtual Machine as the default violated Section 2 because Microsoft had offered no procompetitive justification. Id. Similarly, the court upheld the
finding that Microsoft’s campaign to deceive developers and its threats to Intel to dissuade
Intel from supporting Java both violated Section 2. Id. at 76–78.
211. Id. at 81.
212. Id.
213. Id. at 85 (quoting Kodak I, 504 U.S. 451, 461–62 (1992)); see also Jefferson Parish
Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 17 (1984). (“[T]he Court has held that [if] market
power exists and is being used to restrain competition in a separate market, per se condemnation [of tying is] appropriate.”).
214. Microsoft 2001, 253 F.3d at 84.
215. The court conceded that “we believed our interpretation of the term ‘integrated
product’ [in Microsoft 1998] was consistent with the test for separate products under tying
law,” but pointed out that the Microsoft 1998 opinion had made it clear that the “antitrust
question is of course distinct” and that the conclusion that Internet Explorer and Windows
95 were integrated was subject to reexamination on a more complete record. Id. at 92 (internal quotation marks omitted). The Microsoft 2001 opinion did, however, indicate that any
disclaimer in Microsoft 1998 of judicial competence to evaluate product design did not
conform to prevailing antitrust doctrine. Id.
356
Harvard Journal of Law & Technology
[Vol. 16
that Windows and Internet Explorer were, in fact, two separate products.216 Nonetheless the court concluded that “applying per se analysis
to such an amalgamation creates undue risk of error and of deterring
welfare-enhancing innovation.”217 Accordingly, it remanded the case
to the district court for its evaluation of whether, under the Rule of
Reason, the procompetitive advantages of integration outweighed the
costs to consumers of impairing consumers’ ability to make direct
price/quality tradeoffs in the tied market.218
In short, it seems fair to say that the D.C. Circuit abandoned in
Microsoft 2001 its suggestion in Microsoft 1998 that integrated products are not separate products for purposes of assigning tying liability.
The court also demonstrated its willingness and competence to dig
into the technology of operating systems and other software to determine which products are separate. At the same time, the Microsoft
2001 decision could lead to the same end result. In particular, asking
whether the procompetitive advantages of integration outweigh the
anticompetitive restrictions on consumer choice may turn out to be
similar to asking whether there are functionalities or advantages available from the bundle that are not available if the two products are
bought separately but used together.
Ultimately, the D.C. Circuit partially affirmed the district court's
antitrust violation findings but vacated the remedy of divestiture.219
On remand, the district court was required to fashion a remedy appropriately tailored to the revised liability findings.220 On November 1,
2002, the court entered the final consent decree.221
E. Other Relevant Cases
Discussion of several additional patent and copyright cases is
necessary to demonstrate the full range of decisions in this area. As
above, the courts struggle in these cases to clearly and consistently
define the extent of power that intellectual property law gives patent
and copyright holders in the face of antitrust law. Also as above, each
court approaches this dilemma uniquely, in both interpretation and
application of intellectual property and antitrust law, and thus courts
come to distinct and sometimes conflicting conclusions. This, of
course, only adds to the complexity in this field, and increases the
need for a simpler, comprehensive solution.
216. Id. at 89.
217. Id. at 89–90.
218. Id. at 94.
219. Id. at 105.
220. Id.
221. Microsoft 2002, 231 F. Supp. 2d 144 (D.D.C. 2002).
No. 2]
Adverse Possession for Intellectual Property
357
1. Other Patent Cases
a. Intergraph v. Intel
Intergraph Corp. v. Intel Corp.222 involved Intel’s threat to discontinue providing advanced technical information to a microprocessor customer that had sued Intel for patent infringement. Intel
manufactured high performance microprocessors that were sold to a
variety of OEMs. One such OEM, Intergraph, made workstations for
computer-aided graphics. From 1987 to 1993, Intergraph based its
workstations on its own microprocessor technology called “Clipper,”
but in 1993 Intergraph changed over to Intel microprocessors.223 In
1994, Intel designated Intergraph a “strategic customer” and began to
provide special benefits such as early access to technology under nondisclosure agreements.224
In 1996, Intergraph determined that certain OEMs were infringing
several of its Clipper patents through their use of Intel microprocessors.225 Intel attempted to negotiate access to the Clipper technology,
but unsuccessfully. Intergraph sued Intel for patent infringement in
1997.226 Intel responded by threatening to cut off the special strategic
customer benefits, and Intergraph sought to enjoin Intel from doing
so. The district court characterized Intel’s refusal to deal as denying
access to an essential facility227 and found that as such this refusal
violated Section 2 of the Sherman Act.228
In an opinion issued two years before Xerox, the Federal Circuit
reversed, holding that unilateral conduct permitted under patent and
copyright laws is not subject to antitrust scrutiny.229 The court also
held that “Intel’s conduct with respect to Intergraph does not constitute the offense of monopolization or the threat thereof in any market
relevant to competition with Intergraph.”230 Thus, absent direct competition between Intel and Intergraph, there could be no essential facility for antitrust purposes.231
222. 195 F.3d 1346 (Fed. Cir. 1999).
223. See id. at 1350.
224. See id.
225. See id.
226. Id.
227. Intergraph Corp. v. Intel Corp., 3 F. Supp. 2d 1255, 1278 (N.D. Ala. 1998).
228. Id.
229. Intergraph, 195 F.3d at 1354–55 (“Unilateral conduct that may adversely affect another’s business situation, but is not intended to monopolize that business, does not violate
the Sherman Act.”).
230. Id. at 1356.
231. Id. at 1357–58 (“The notion that withholding of technical information and samples
of pre-release chips violates the Sherman Act, based on essential facility jurisprudence, is an
unwarranted extension of precedent and can not be supported on the premises presented.
The district court erred in holding that Intel's superior microprocessor product and Inter-
358
Harvard Journal of Law & Technology
[Vol. 16
Although Intel prevailed before the Federal Circuit, a concurrent
FTC action ultimately forced Intel to acquiesce.232 Per the consent
decree, Intel agreed to continue providing advanced technical information for ten years to microprocessor customers even if one of those
customers initiated an infringement action against Intel.233
b. SCM Corp. v. Xerox Corp.
Another relevant case involving antitrust liability for actions
within the scope of the patent grant is SCM Corp. v. Xerox Corp.,234
which the Ninth Circuit cited several times in Kodak II.235 As part of
its strategy to develop the plain-paper copier, “Xerox had more than
1,000 patents, mostly related to xerography and plain-paper copying.”236 Xerox only used 35 to 40 percent of those patents in actual
Xerox products.237 In essence, Xerox had accumulated most of those
patents to create a defensive thicket around its photocopier technology.238 Having made substantial R&D investments, Xerox “felt that it
wanted to exploit its copier technology on its own.”239 “Therefore,
throughout the 1960s and into the early 1970s, Xerox refused to grant
licenses for plain paper copying under its patents [but] did grant licenses under its patents for other fields, including coated paper copiers.”240 SCM had obtained a license for coated paper copying patents
but then attempted unsuccessfully to obtain a license for plain paper
copying patents.241
SCM filed an antitrust claim against Xerox alleging that “Xerox’s
acquisition of its patents and subsequent exercise of the exclusionary
power in them violated the antitrust laws and injured SCM.”242 SCM
asserted that Xerox acquired most of its patents to block others from
making plain paper copiers, rather than for protecting improved Xerox
graph’s dependency thereon converted Intel’s special customer benefits into an ‘essential
facility’ under the Sherman Act.”).
232. See In the Matter of Intel Corp., F.T.C. Docket No. 9288, http://www.ftc.gov/
os/1998/9806/intelfin.cmp.htm. The complaint alleged not only that Intel’s actions regarding Intergraph were illegal, but also that Intel was engaging in similar conduct against Digital and Compaq.
233. See Agreement Containing Consent Order, In the Matter of Intel Corp. (filed Mar.
17, 1999), http://www.ftc.gov/os/1999/9903/d09288intelagreement.htm.
234. 645 F.2d 1195 (2d Cir. 1981).
235. Kodak II, 125 F.3d 1195, 1214, 1216, 1219.
236. Gerald Sobel, The Antitrust Interface with Patents and Innovation: Acquisition of
Patents, Improvement Patents and Grant-Backs, Non-Use, Fraud on the Patent Office,
Development of New Products and Joint Research, 53 ANTITRUST L.J. 681, 685 (1984).
237. See id.
238. See Kurt M. Saunders, Patent Nonuse and the Role of Public Interest as a Deterrent
to Technology Suppression, 15 HARV. J.L. & TECH 389, 393 (2002).
239. Sobel, supra note 236, at 685.
240. Id.
241. See SCM v. Xerox, 645 F.2d at 1200.
242. Id. at 1203.
No. 2]
Adverse Possession for Intellectual Property
359
products and processes, and that Xerox’s failure to use the majority of
its patents “reflected this intent.”243 Although “Xerox’s patents were
so numerous and complex that they created a ‘thicket’ that prevented
[SCM from] designing around the patents,”244 the Second Circuit ultimately held “that where a patent has been lawfully acquired, subsequent conduct permissible under the patent laws cannot trigger any
liability under the antitrust laws.”245
2. Other Copyright Cases
a. Sega Enterprises v. Accolade
In the first case, Sega Enterprises v. Accolade Inc.,246 Sega manufactured video game consoles and cartridges for home entertainment
use.247 Accolade was a competitor in the video game cartridge market
but did not manufacture its own game consoles.248 Sega had developed a system to protect its trademark rights — the Licensed Trademark Security System (“TMSS”) — by which the Sega console read a
game cartridge for specific computer code.249 If the game cartridge
included the computer code, the console would display the message
“PRODUCED BY OR UNDER LICENSE FROM SEGA.”250 If the
TMSS code was not found, the game cartridge would not operate in
the console.251
Using reverse engineering, Accolade analyzed Sega’s game cartridges to determine which code was necessary for the cartridges to be
compatible with the Sega console, and then used that code to develop
their own game cartridges for play on the console.252 Because Accolade also copied the TMSS code, however, Accolade’s cartridges also
prompted the message that Accolade games are “PRODUCED BY
OR UNDER LICENSE FROM SEGA,” even though Accolade had no
license from Sega to produce compatible game cartridges.253
Sega filed suit alleging copyright infringement, among other
things, and sought to enjoin Accolade.254 The district court granted the
injunction,255 but the Ninth Circuit reversed,256 applying the doctrine
243. Sobel, supra note 236, at 685.
244. Id. (quoting SCM v. Xerox, 645 F.2d at 1203).
245. SCM v. Xerox, 645 F.2d at 1206.
246. 977 F.2d 1510 (9th Cir. 1992).
247. Id. at 1514.
248. Id.
249. Id. at 1515.
250. Id.
251. Id. at 1515–1516.
252. Id. at 1516.
253. Id.
254. Id.
255. Id.
256. Id. at 1532.
360
Harvard Journal of Law & Technology
[Vol. 16
of fair use.257 The Ninth Circuit held that where reverse engineering
“is the only way to gain access to the ideas and functional elements
embodied in a copyrighted computer program and where there is a
legitimate reason for seeking such access, disassembly is a fair use of
the copyrighted work, as a matter of law.”258
b. MAI Systems v. Peak Computer
In the second case, MAI Systems Corp. v. Peak Computer Inc.,259
MAI was a computer manufacturer that designed software to run on
its computers.260 MAI also provided service for its customers’ computers and the software necessary to operate its computers, including
the operating system.261 Peak Computer maintained computer systems
for its clients, including more than one hundred clients that used MAI
computers.262 Peak competed with MAI in servicing MAI equipment,
which comprised between 50 percent and 70 percent of Peak’s business.263 As part of its service program, Peak technicians would often
“operate the computer and its operating system software in order to
service the machine.”264 In 1992, MAI sued Peak for copyright infringement, among other things, alleging that every time a Peak technician operated the equipment or loaded a diagnostic program, an
infringing copy of the software was created in the computer’s
RAM.265 The Ninth Circuit held that “the loading of copyrighted
software into RAM creates a ‘copy’ of that software in violation of
the Copyright Act,”266 and that such copying by Peak was unauthorized,267 even if for purely diagnostic purposes as part of servicing or
repairing the customer’s equipment.268
c. Data General v. Grumman
The last case we discuss in this section is Data General Corp. v.
Grumman Corp.,269 which the courts in both Kodak II 270 and Xerox271
257. See discussion of fair use supra note 20 and accompanying text.
258. Sega Enterprises, 977 F.2d at 1527–28 (citing the disassembly of the object code
reverse engineering technique).
259. 991 F.2d 511 (9th Cir. 1993).
260. See id. at 513.
261. See id.
262. See id.
263. See id.
264. See id.
265. Id. at 517.
266. Id. at 518.
267. Id. at 517.
268. See id. at 518, n. 4 and accompanying text.
269. 36 F.3d 1147 (1st Cir. 1994).
270. Kodak II, 125 F.3d at 1211, 1215–16, 1218–19.
271. Xerox, 203 F.3d 1322, 1327–28 (Fed. Cir. 2000).
No. 2]
Adverse Possession for Intellectual Property
361
cited.272 Data General and Grumman competed in the service market
for computers manufactured by Data General.273 Although Data General had “no more than a 5 percent share of the highly competitive
‘primary market’ for mini-computers, [it controlled] approximately 90
percent of the ‘aftermarket’ for service of [Data General] computers.”274 Grumman was the leading ISO275 with approximately 3
percent of the available service business.276
The first ISOs appeared in the early 1970s,277 and from 1976 until
some point in the mid-1980s, Data General fostered the growth of
ISOs with relatively liberal policies concerning ISO access to service
tools.278 Data General sold or licensed diagnostics directly to ISOs,
and allowed them to use diagnostics that had been sold or licensed to
Data General equipment owners and to obtain spare parts manufactured by Data General or other manufacturers.279 Data General “allowed (or at least tolerated) requests by [ISOs for Data General’s]
repair depot to fix malfunctioning circuit boards, the heart of a computer's central processing unit (“CPU”).”280 Some ISOs could even
purchase engineering change order kits, classroom training, schematics, and other documentation from Data General.281 Grumman argued
“that [Data General’s] liberal policies were beneficial to [Data General] because increased capacity (and perhaps competition) in the service aftermarket would be a selling point for [Data General]
equipment.”282
Much as Kodak and Xerox changed their policies to maximize
revenue from their service business, so too did Data General. In the
mid-1980s, Data General began to refuse to provide many service
tools directly to ISOs, allow them to use the Data General repair depot, permit ISOs to attend training classes, or to purchase schematics,
documentation, change order kits, and certain spare parts.283 Finally,
Data General “severely restricted the licensing of ADEX, a new software diagnostic for its MV computers. The MV series was at once
[Data General’s] most advanced computer hardware and an increas-
272. Hovenkamp, Janis, and Lemley also suggest that Data General should be read together with Kodak I, II and Xerox. See IP AND ANTITRUST, supra note 1, § 13.3d.
273. See Data General, 36 F.3d at 1152.
274. Id.
275. The term used in the Data General opinion is “TPM” (Third Party Maintainer), but
we will use ISO for the sake of consistency with Xerox and both Kodak opinions.
276. See Data General, 36 F.3d at 1152.
277. See id. at 1153.
278. See id. at 1154.
279. See id.
280. See id.
281. See id.
282. See id.
283. See id.
362
Harvard Journal of Law & Technology
[Vol. 16
ingly important source of sales and service revenue”284 for Data General.
While the ISOs could obtain parts, service tools, and most required software from other sources,285 access to ADEX was a different matter. Data General refused to license ADEX to its own service
customers or to the customers of ISOs; it was only available to Data
General technicians or those equipment owners that performed their
own service.286 ISOs were also unable to obtain ADEX equivalents
from sources other than Data General.287 ISOs such as Grumman
found various ways to circumvent Data General's ADEX restrictions.
Some former Data General employees absconded with copies of
ADEX when they left Data General and joined Grumman.288 In addition, although Data General required service customers to return copies of ADEX to Data General should they cancel their service
agreement, few customers did so.289 Grumman technicians could also
use and duplicate copies of ADEX left behind by Data General field
engineers. Grumman was thus able to acquire copies of ADEX in order to maintain libraries of diagnostics, so that Grumman technicians
could more easily provide service to customers with Data General's
MV computers.290
Data General sued Grumman for copyright infringement in
1998.291 Grumman asserted that Data General could not maintain a
copyright infringement action because Data General “had used its
ADEX copyrights to violate Sections 1 and 2 of the Sherman Antitrust
Act.”292 The district court granted Data General’s motions for summary judgment with respect to Grumman’s tying claim under Section
1 as well as its monopolization claim under Section 2, and the First
Circuit affirmed both rulings.293 The First Circuit held that Grumman
failed to prove that ADEX and Data General support were two distinct
products, and thus, as a matter of law, no tying arrangement could
exist.294 As to the Section 2 claims, the First Circuit held that “while
exclusionary conduct can include a monopolist’s unilateral refusal to
license a copyright, an author’s desire to exclude others from use of
284. See id.
285. See id.
286. See id.
287. See id. (“At least two other diagnostics designed to service Data General's MV computers may have become available as early as 1989, but no fully functional substitute was
available . . . [as of] 1992.”).
288. Id.
289. Id. at 1154–55.
290. Id. at 1155.
291. Id.
292. Id. at 1156.
293. Id. at 1178. The district court also disagreed with Grumman’s characterization of
ADEX as an essential facility. Data Gen. Corp. v. Grumman Sys. Support Corp., 761 F.
Supp. 185, 192 (D. Mass. 1991).
294. Data General, 36 F.3d at 1181.
No. 2]
Adverse Possession for Intellectual Property
363
its copyrighted work is a presumptively valid business justification for
any immediate harm to consumers.”295
IV. MAKING SENSE OF KODAK, XEROX, AND MICROSOFT
The situation that confronts the intellectual property marketplace
is one fraught with confusion and conflicting directives. The Ninth
Circuit’s approach in Kodak II296 seems deeply flawed.297 By making
liability turn on the defendant’s subjective intent, the Ninth Circuit
would make it difficult or impossible for defendants to have the case
decided on a motion for summary judgment. It basically rewards the
company that has best educated its managers on the need to constantly
repeat in their e-mails and elsewhere that the reason why they are not
dealing with others is because of their desire to exploit their intellectual property rights.298 Another company engaging in essentially the
same conduct would be treated differently not because its conduct was
more or less harmful to competition, but because it had done a poorer
job schooling its managers. Instead, the test should be whether the
practice in fact harms competition.
The D.C. Circuit’s position in Microsoft 2001299 seems equally
problematic. Following the court’s baseball bat analogy, it would
seem that if a fan ran onto the field and approached a player warming
up inside the on-deck circle from behind, the player should not be
liable if the fan were struck by the bat during the normal course of
swinging the bat.300 Yet if one followed the logic of Microsoft 2001,
liability would have to be imposed no matter how legitimate the bat
swinging may have been. Had the D.C. Circuit correctly followed the
logic of the Federal Circuit’s Intergraph301 and Xerox302 opinions, it
would have focused on whether Microsoft’s activity exceeded the
scope of its grant of intellectual property protection. While its baseball
bat analogy carries rhetorical heft, absent an examination of the con-
295. Id. at 1187.
296. Kodak I, 125 F.3d 1195 (9th Cir. 1997), cert. denied, 523 U.S. 1094 (1998).
297. See, e.g., Michael A. Carrier, Unraveling the Patent-Antitrust Paradox, 150 U. PA.
L. REV. 761, 794 (2002) (“There are manifest difficulties in centering analysis on the defendant's intent.”).
298. See id. (“[T]here will always be the concern that documentation of beneficent intent
will be made with an eye toward future litigation.”).
299. Microsoft 2001, 253 F.3d 34, 62 (D.C. Cir. 2001).
300. Responding to Microsoft’s assertion that if intellectual property rights have been
lawfully acquired then their subsequent exercise cannot give rise to antitrust liability, the
court stated “[t]hat is no more correct than the proposition that use of one’s personal property, such as a baseball bat, cannot give rise to tort liability.” Microsoft 2001, 253 F.3d 34,
62 (D.C. Cir. 2001).
301. Intergraph Corp v. Intel Corp, 195 F.3d 1346 (Fed. Cir. 1999).
302. Xerox, 203 F.3d 1322 (Fed. Cir. 2000).
364
Harvard Journal of Law & Technology
[Vol. 16
text of batsmanship, the analogy ends up like a corked bat that has
shattered at the plate, exposed as hollow for all the world to see.303
While Microsoft 2001 dealt only with copyrighted products, Kodak II dealt with both patented and copyrighted products. Since Nobelpharma,304 which was decided less than one year after Kodak II,
however, all cases where patent law conflicts with other causes of
action, including antitrust, have been handled by the Federal Circuit.305
In Xerox, the Federal Circuit explicitly rejected the intent test of
Kodak II, so for patents, at least, one could argue that Kodak II is no
longer valid law. Instead, Xerox holds that any activity within the
scope of the patent grant is fair game. We see no reason why the ability to refuse to license intellectual property should turn on whether the
asset is protected by copyright or by patent. Such an approach would
be particularly troublesome as applied to software code and other
products eligible for both forms of protection.
While we are most troubled by the treatment of intellectual property in Kodak II and Microsoft 2001, we also feel that Xerox goes too
far in that it allows Xerox to change its policies retroactively. Had the
Federal Circuit limited Xerox to the implementation of new policies
for new products rather than also allowing retroactive policy changes
for existing products, the resulting opinion would have avoided the
anticompetitive problems that the Supreme Court identified in Aspen
Skiing,306 and which courts have admonished against in franchise tying cases.307
While granting absolute immunity within the scope of the patent
grant allows for quick disposal of cases at summary judgment, the
subjective intent test in Kodak II will almost always have to be decided by a jury. An objective test that can be handled at summary
judgment is desirable, but the Xerox position goes too far. Additionally, what is needed is an objective test that is based on market realities rather than “[l]egal presumptions that rest on formalistic
distinctions.”308 We propose such a test in the following section.
303. To cork a baseball bat, drill a hole 12 to 14 inches down into the barrel and fill it
with ground cork. “A heavy bat is thus able to be swung with the quickness of a bat several
ounces lighter, which can cause a ball to carry much farther than it otherwise would have.”
Kevin Paul Dupont, Cheating in Other Sports? BOSTON GLOBE, Apr. 17, 1998, at C7.
304. Nobelpharma AB v. Implant Innovations, Inc., 141 F.3d 1059 (Fed. Cir. 1998)
305. See id. But see discussion of Holmes Group v. Vornado Air Circulation Sys., 535
U.S. 826 (2002), discussed in supra note 157.
306. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985). See supra
text accompanying notes 56–58.
307. See supra notes 80–83 and accompanying text.
308. Kodak I, 504 U.S. 451, 466 (1992).
No. 2]
Adverse Possession for Intellectual Property
365
A. Our Proposal
Our proposal involves a refinement of the Federal Circuit’s position that any activity within the scope of the intellectual property grant
is permissible. We accomplish that refinement by incorporating a
variation on the concept of adverse possession for the intellectual
property space. Just as adverse possession has potential benefits in the
arena of real property, similar benefits could be obtained by applying
a variant of this principle to intellectual property. Of course, analogizing a real property concept in the intellectual property space is never a
perfect fit. The nature of intellectual property is such that almost any
real property analogy will require modification. Unlike real property,
intellectual property is an intangible good and is therefore nonrivalrous.309 Intellectual property is also non-excludable,310 and thus
difficult to monitor and control in terms of dissemination and use
without additional legal protections.311
Alternative analogies may also exist that are preferable on certain
dimensions yet are also inadequate on other dimensions. An obvious
alternative to the adverse possession analogy would be an easement
for intellectual property, but for reasons discussed in Part V, we reject
the easement analogy and proceed with the adverse possession analogy.
Given that intellectual property owners should be rightfully entitled to claim the full scope of their property grant, we agree with the
Federal Circuit’s position that they should be able to exclude other
firms from developing ancillary markets within the scope of the original intellectual property grant, via unilateral refusals to license their
intellectual property. If, however, other firms are not excluded and the
ancillary markets develop and provide a service that is separately demanded by consumers, then we propose that after a sufficient period
of time, adverse possession would preclude the original intellectual
property owners from excluding the other firms from the ancillary
market based on existing products. New markets based on new products would not be affected, however, unless (1) the intellectual prop309. A non-rivalrous good is one that can be used simultaneously by more than one person without diminishing the utility of its use. See Paul Romer, Endogenous Technological
Change, 98 J. POL. ECON. S71, S74 (Oct. 1990) (defining a non-rivalrous good as having
“the property that its use by one firm or person in no way limits its use by another”); see
also Lawrence Lessig, The Law of the Horse: What Cyberlaw Might Teach, 113 HARV. L.
REV. 501, 526 (1999) (discussing the non-rivalrous nature of intangible property).
310. See Romer at S74 (“A good is excludable if the owner can prevent others from using
it.”).
311. Id. (“A good such as the code for a computer program can be made excludable by
means of a legal system that prohibits copying”). Even with additional legal protections,
detection of unauthorized use or dissemination may prove difficult. To enhance excludability, Romer also suggests developing a “means of encryption [or other] copy protection
schemes.” Id.
366
Harvard Journal of Law & Technology
[Vol. 16
erty owner repeats the same process of failing to claim the entire
scope of the intellectual property grant relative to potential ancillary
markets, or (2) the intellectual property owner has market power and
the underlying product is one that is characterized by network effects.
In this latter (and probably rare) situation, we propose an application
of the Rule of Reason if the underlying product is found to be an essential facility and the holder of the essential facility competes in the
ancillary market.312 Before we explore the specific application of our
proposal, however, it is instructive to review some of the justifications
for adverse possession as applied to real property.
1. Relevant Justifications for Adverse Possession
The most relevant economic justification for adverse possession
is that “adverse possessors who have occupied a piece of land for a
long period may have developed considerable reliance interests that
would be lost if the true owner could reclaim title at any time.”313
Judge Posner characterized the rationale as follows:
Over time, a person becomes attached to property
that he regards as his own, and the deprivation of the
property would be wrenching. Over the same time, a
person loses attachment to property that he regards
as no longer his own, and the restoration of the property would cause only moderate pleasure. This is a
point about diminishing marginal utility of income.
The adverse possessor would experience the deprivation of the property as a diminution of his wealth; the
original owner would experience the restoration of
the property as an increase in his wealth. If they have
the same wealth, then probably their combined utility
will be greater if the adverse possessor is allowed to
keep the property.314
Other scholars have characterized this justification as moral in nature.315 Professor Singer described the situation this way:
312. See supra Part II.D.A for discussion of the justifications for the Essential Facilities
Doctrine.
313. Thomas J. Miceli & C. F. Sirmans, An Economic Theory of Adverse Possession, 15
INT’L REV. L. & ECON. 161 (1995); see also Jeffrey M. Netter et al., An Economic Analysis
of Adverse Possession Statutes, 6 INT. REV. L. & ECON. 217 (1986).
314. RICHARD POSNER, ECONOMIC ANALYSIS OF LAW §3.11, at 89–90 (5th ed. 1998).
315. See, e.g., Joseph William Singer, The Reliance Interest in Property, 40 STAN. L.
REV. 611, 667 (1988) (“It is morally wrong for the true owner to allow a relationship of
dependence to be established and then to cut off the dependent party.”); see also Thomas W.
Merrill, Property Rules, Liability Rules, and Adverse Possession, 79 NW. U. L. REV. 1122
(1984) (discussing adverse possession in moral terms).
No. 2]
Adverse Possession for Intellectual Property
367
The true owner and the adverse possessor [develop]
a kind of relationship. The possessor comes to expect
and may have come to rely on the fact that the true
owner will not interfere with the possessor's use of
the property. If the adverse possessor were to be
ousted from the property, she would experience a
loss. The adverse possessor's interests grow stronger
over time as she develops legitimate expectations
that the true owner will continue to allow her to control the property. In other words, she may rely on
continued access, both in the sense of relying on use
of the land itself and relying on the relationship that
makes such access possible.
. . . The possessor has come to expect continued access to the property, and the true owner has fed those
expectations by her actions (or her failure to act). It
is morally wrong for the true owner to allow a relationship of dependence to be established and then to
cut off the dependent party.
. . . After the true owner has acquiesced in the adverse use for a long enough period of time, . . . [t]he
rules protect the more vulnerable party to the relationship by shifting ownership from the true owner
to the adverse possessor.316
This situation seems quite analogous to the fact patterns in both
Kodak and Xerox. If we view the maximum scope of the patent grant
as a piece of property, the servicing of products based on the patent
grant is initially within the metes and bounds of the property grant.
Kodak and Xerox initially left part of that space vacant, and the ISOs
subsequently occupied that space.317
Given that antitrust is designed to enhance consumer welfare by
protecting competition rather than competitors,318 a reliance interest
by the ISOs alone would be insufficient.319 Kodak and Xerox custom316. Singer, supra note 315, at 666–67.
317. At first glance, the patent law doctrine of implied license might appear to be a valid
alternative, but the specific requirements for its application is not present in the situations
discussed here. An implied license can be imposed in situations where the patent owner has
acquiesced in a situation of infringement. See Wang Laboratories v. Mitsubishi Electronics,
103 F.3d 1571, 1580 (Fed. Cir. 1997). Under our proposal, however, infringement never
gives rise to adverse possession, and thus this doctrine is inapplicable
318. See, e.g., Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 488 (1977);
Brown Shoe Co. v. United States, 370 U.S. 294, 320 (1962).
319. A social welfare analysis, however, that is agnostic as to the intent of the antitrust
regime would also examine the investment of the ISOs in terms of such costs, separate and
368
Harvard Journal of Law & Technology
[Vol. 16
ers also have a reliance interest, however. When they were purchasing
their equipment, the existence of ISO service may have been factored
into their purchase decision. It is the reliance interest of the consumers
that provides the justification for the application of adverse possession
in the intellectual property space.320 As a result, the right to limit the
IP owner’s right to exclude rests with the existing customers and is
not specific to any one ISO. Once that customer reliance interest is
created, any ISO, regardless of when it entered the market, can assert
it on behalf of the existing customers.
A different economic justification for adverse possession “is that
it tends to prevent valuable resources from being left idle for long
periods of time by specifying procedures for a productive user to take
title from an unproductive user.”321 This Lockean notion of highest
and best use of the land,322 which rewards “the useful laborer at the
expense of the sluggard,”323 would suggest that if the ISOs are able to
deliver a superior level of service, they should be allowed to continue
to do so if Kodak and Xerox did not exclude them from the service
market from the beginning.
Our proposal of adverse possession for intellectual property only
applies if the adverse possession occurs in a marketplace in which the
adverse possessor is already established. If the ISO service were inferior such that consumers did not demand it, then under the separate
products test of Jefferson Parish,324 there would be no separate market
and thus no anticompetitive conduct. In Jefferson Parish, the Court
stated that “no tying arrangement can exist unless there is a sufficient
demand for the purchase of [service A] separate from [product B] to
identify a distinct product market in which it is efficient to offer [serapart from any customer based reliance interest. Allowing ISOs to invest heavily in product
specific training or equipment and then forcing them to abandon those investments would
seem deleterious to overall economic welfare. We revisit this issue of sunk investments later
in this Part. See infra note 352 and accompanying text.
320. We might also make a similar argument under a theory of detrimental reliance or
breach of the implied covenant of good faith and fair dealing from a contract law perspective. However, the adverse possession concept seems to be less of a conceptual stretch.
Neither Kodak nor Xerox ever explicitly promised that ISO service would be available to
their customers, nor did they agree to supply parts to ISOs in perpetuity. We are also unaware of an instance where a reliance interest for one party (the customer) was sustainable
based on a separate reliance interest of a third party (the ISOs). Thus, we do not pursue
detrimental reliance under contract law any further.
321. ROBERT COOTER & THOMAS ULEN, LAW AND ECONOMICS 156 (1988); see also
Merrill, supra note 315, at 1130 (describing adverse possession as “a social policy favoring
‘active’ owners of property, who develop or exploit their land, rather than ‘passive’ owners”).
322. See generally John Locke, Second Treatise of Government § 25, in TWO TREATISES
OF GOVERNMENT 336 (P. Laslett rev. ed. 1963) (1690).
323. Carol M. Rose, Possession as the Origin of Property, 52 U. CHI. L. REV. 73, 79
(1985).
324. Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 21 (1984) (“a tying arrangement cannot exist unless two separate product markets have been linked”).
No. 2]
Adverse Possession for Intellectual Property
369
vice A] separately from [product B].”325 Thus, adverse possession for
a portion of the intellectual property space can be justified based on
the reliance interest of the customers of the patented products as well
as the increased customer welfare of allowing the competitive service
to continue to be offered.326
Adverse possession for real property is far from instantaneous and
is difficult to accomplish. To prevail on a claim for adverse possession, the adverse possessor must demonstrate (a) actual possession
that is (b) open and notorious, (c) exclusive, (d) continuous and (e)
adverse or hostile (f) for the statutory period.327 Any application of
adverse possession to intellectual property should be similarly constrained; however, the nature of intellectual property necessitates
some modification to the doctrine. Actual possession would be established by one or more firms occupying the ancillary market within the
scope of the patent grant. This modification would also satisfy the
open and notorious requirement, given that the ISO-type firms would
be purchasing or licensing with the full knowledge of the intellectual
property owner.
Exclusive occupation would differ from the real property concept
because of inherent differences between real property and intellectual
property. It is physically impossible for a sufficiently small piece of
real property to be simultaneously occupied by more than one person,
yet those same two people could simultaneously make use of the same
piece of intellectual property.328 In the intellectual property arena, the
determining factor would be whether the intellectual property owner
(or an exclusive licensee) occupies the ancillary market exclusively.
In essence, nonexclusivity on the part of the intellectual property
owner would be the required criteria for a successful claim for adverse
possession, as long as the additional occupiers were non-infringing.
We should make it clear that infringement of the intellectual property
owner’s IP rights would never qualify for adverse possession.
While the concept of continuity would be similar, the requirement
that the possession be adverse or hostile would be different. Given
that the intellectual property owner is the source for patented spare
parts, one could argue that the occupation of the ancillary market is
permissive. Since permissive use of real property generally defeats a
claim for adverse possession,329 we would therefore propose that the
325. Id. at 21–22.
326. There are of course other justifications for adverse possession. A particularly interesting justification based on modern experimental psychology and the theory of loss aversion can be found in Jeffrey Evans Stake, The Uneasy Case for Adverse Possession, 89
GEO. L.J. 2419 (2001).
327. JOSEPH WILLIAM SINGER, INTRODUCTION TO PROPERTY LAW 137 (2001).
328. This distinction highlights the difference between the rivalrous nature of real property and the non-rivalrous nature of intellectual property.
329. See SINGER, supra note 315, at 142. But see Peters v. Juneau-Douglas Girl Scout
Council, 519 P.2d 826 (Alaska 1974) (possession by Alaskan Tlingit Indian elder was held
370
Harvard Journal of Law & Technology
[Vol. 16
occupation of the ancillary market must only be competitive to qualify
as adverse or hostile. If the intellectual property owner chooses not to
compete, the potential for competition would also be sufficient for a
claim of adverse possession. Again, while infringement might be
viewed as an adverse or hostile activity, we do not suggest that such
infringement ever qualify for adverse possession.
In terms of hostile or adverse occupation, our proposal for the intellectual property space clearly differs from traditional adverse possession for real property. Our construction, however, closely follows
Singer’s reinterpretation of adverse possession, which suggests that
the doctrine requires only longstanding use rather than requiring nonpermissive longstanding use.330 We propose that as long as the use is
non-infringing, the adverse possession doctrine for the intellectual
property space should not necessarily distinguish between longstanding permissive use and longstanding nonpermissive use.
The last requirement for a successful adverse possession claim is
that the occupation be for a sufficiently long period of time. Given
that intellectual property grants are limited in duration, it would seem
to follow that its required timeframe would be shorter than the required timeframe for real property. Although we do not suggest what
the appropriate period of time should be for intellectual property, the
Data General331 case discussed earlier might provide a basis for determining a sufficient period.332 As Hovenkamp, Janis, and Lemley
point out, the First Circuit held that because “Data General’s old policy of permitting ISO competition had never led to a competitive market, . . . the withdrawal of its support could not be proof of
anticompetitive effect.”333 Thus, a sufficient period of time might be
stated in terms of market competitiveness rather than in months or
years, which would also be consistent with our requirement that the
occupation be competitive to qualify as hostile or adverse. Alternatively, the sufficient time period could be stated in terms of the time
necessary to make sufficient investment to compete in the ancillary
market. Such a determination would take into account such factors as
the level of investment required or the asset specificity of such investment.
In addition to a theoretical justification for our proposal, cases
from several circuits provide additional support for the idea that a
to be sufficiently hostile under an objective test of hostility to support adverse possession
even though plaintiff’s possession of the property was in fact permissive).
330. See Singer, supra note 315, at 666–67. Note that Singer does not suggest that the
distinction between permission and nonpermission is irrelevant. He posits instead that such
a distinction should be a factor in determining whether the reliance interest should be protected or not. See id.
331. Data Gen. Corp v. Grumman Corp., 761 F. Supp. 185 (D. Mass. 1991).
332. See supra Part III.E.3.
333. IP AND ANTITRUST, supra note 1, at 13–26.
No. 2]
Adverse Possession for Intellectual Property
371
change in longstanding business practices should be the threshold
question from an antitrust standpoint. Both the First334 and the Seventh335 Circuits have limited Kodak I336 to situations in which the
manufacturer’s policy was not generally known. The situation before
the Sixth Circuit in PSI Repair Services. v. Honeywell, Inc.,337 seems
the closest to both the situations in Kodak and Xerox, and the court’s
reasoning seems quite consistent with our proposal.
Honeywell manufactured and sold industrial control equipment
that contained and depended heavily on printed circuit boards.338
Honeywell did not repair the boards when they failed, but instead replaced them and charged the customer for the replacement, “so long
as the customer return[ed] the defective board.”339 Although companies such as PSI offered circuit-board repair services to owners of
industrial control equipment, “Honeywell has a stated policy of not
selling its components to anyone.”340 Thus, there was no ISO market
for Honeywell equipment, because “Honeywell’s own actions . . .
essentially limited the existence of a separate market for components.”341 In distinguishing Kodak I, the Sixth Circuit found merit in
Honeywell’s contention “that that while Kodak changed its service
and parts policy to include the tie after many Kodak customers purchased their equipment, Honeywell has consistently maintained its
policy of board replacement.”342 In essence, “the Kodak [I] Court's
finding of a material issue of fact on the definition of the relevant
market resulted from Kodak’s change in policy.”343 Kodak I did not
explicitly state the extent that the change in Kodak’s policy “affected
the Court’s analysis. Nonetheless, it suggested that the outcome might
have been different had Kodak presented evidence that its restrictive
334. See Lee v. Life Ins. Co. of N. Am., 23 F.3d 14, 20 (1st Cir. 1994), cert. denied, 513
U.S. 964 (1994) (“The timing of the ‘lock in’ at issue in Kodak was central to the Supreme
Court's decision . . . . Had previous customers known, at the time they bought their Kodak
copiers, that Kodak would implement its restrictive parts-servicing policy, Kodak’s ‘market
power,’ i.e., its leverage to induce customers to purchase Kodak servicing, could only have
been as significant as its [market power] in the copier market, which was stipulated to be
inconsequential or nonexistent.”).
335. See Digital Equip. Corp. v. Uniq Digital Techs., Inc., 73 F.3d 756, 763 (7th Cir.
1996) (“The Court did not doubt in Kodak that if spare parts had been bundled with Kodak’s
copiers from the outset, or Kodak had informed customers about its policies before they
bought its machines, purchasers could have shopped around for competitive life-cycle
prices. The material dispute that called for a trial was whether the change in policy enabled
Kodak to extract supra-competitive prices from customers who had already purchased its
machines.”).
336. Kodak I, 504 U.S. 451 (1992).
337. 104 F.3d 811 (6th Cir. 1997).
338. See id. at 813.
339. Id.
340. Id.
341. Id. at 816.
342. Id. at 819.
343. Id.
372
Harvard Journal of Law & Technology
[Vol. 16
parts policy was consistently maintained and generally known.”344
The Sixth Circuit stated “that the change in policy in Kodak [I] was
the crucial factor in the Court’s decision. By changing its policy after
its customers were ‘locked in,’ Kodak took advantage of the fact that
its customers lacked the information to anticipate this change.”345
Thus, “it was Kodak’s own actions that increased its customers’ information costs. In our view, this was the evil condemned by the
Court and the reason for the Court’s extensive discussion of information costs.”346
One of the Sixth Circuit’s main holdings, which strongly supports
our proposal, declared that
[i]n light of our reading of Jefferson Parish and Kodak, we thus hold that an antitrust plaintiff cannot
succeed on a Kodak-type theory when the defendant
has not changed its policy after locking-in some of
its customers, and the defendant has been otherwise
forthcoming about its pricing structure and service
policies.347
Because there were “no allegations that Honeywell changed its
parts-restrictive policy in order to lock-in customers, nor has PSI al-
344. Id. The Sixth Circuit noted the majority’s response to Justice Scalia’s dissenting position that “but for Kodak's change in policy, the Court would be faced with a traditional tie
between copiers and aftermarket service.” The majority responded:
The dissent disagrees [with our conclusion in this case] based on its
hypothetical case of a tie between equipment and service. “The only
thing lacking” to bring this case within the hypothetical case, states
the dissent, “is concrete evidence that the restrictive parts policy
was . . . generally known.” But the dissent’s “only thing lacking” is
the crucial thing lacking — evidence. Whether a tie between parts
and service should be treated identically to a tie between equipment
and service, as the dissent and Kodak argue, depends on whether the
equipment market prevents the exertion of market power in the parts
market. Far from being “anomalous,” requiring Kodak to provide
evidence on this factual question is completely consistent with our
prior precedent.
Kodak I, 504 U.S. at 477 n.24 (internal citations omitted). This passage can be read to imply
that had Kodak presented undisputed evidence that it never changed its policy and that its
policy was generally known, the Court would have considered Kodak copiers as the tying
product and service and parts combined as the product being tied. Id.; see also 10 PHILLIP E.
AREEDA, ET AL., ANTITRUST LAW 150 (1996) (“The majority [in Kodak] apparently agreed
with [the dissent] when it emphasized the absence of evidence that Kodak’s policy was
generally known.”); id. at 157 (“The Kodak majority indicated that it would assess the defendant’s power in the interbrand machine market (where it had none) were it ‘generally
known’ to machine buyers that the defendant supplied unique repair parts only in connection with its service, notwithstanding ignorance of lifecycle prices.”).
345. PSI Repair Servs., 104 F.3d at 820.
346. Id.
347. Id.
No. 2]
Adverse Possession for Intellectual Property
373
leged that Honeywell’s policy was not generally known,”348 PSI could
not “establish that Honeywell’s practices in the parts and services
markets [had] contributed to Honeywell’s power in the market for
industrial control equipment,”349 and thus PSI could not prevail.
Having found judicial precedent to bolster the theoretical justification basis for subjecting part of the intellectual property grant to
adverse possession, we extend this concept in the next two sections
first to firms without market power and then to firms with market
power.
2. Firms Without Market Power
If a firm without market power in any relevant product market introduces a new product based in whole or in part on a collection of
intellectual property rights, it would have three options for an ancillary market that would implicate its intellectual property rights (such
as the sale of patented spare parts). First, it could prohibit the development of a secondary market at the outset by refusing to license or
sell to anyone. For example, Apple Computer pursued this strategy
when it refused to license its Mac operating system to hardware
manufacturers interested in selling Apple clones (except for a brief
period when it gave short-term licenses to select hardware manufacturers).350 Because Apple did not have market power in either the personal computer operating system or hardware markets, this tying of
the Mac OS to Apple hardware was permissible under standard tying
analysis, even though the OS and hardware are separate products. Our
proposal would lead to the same result.
A second alternative would be to grant limited rights to the ancillary market using contracts of relatively short duration.351 This second
situation would be analogous to a lease of real property, with ISO investment being analogous to leasehold improvements.352 The third
alternative would be to abstain from excluding other firms in the ancillary market, thus allowing that market to arise naturally.
It is in this third scenario that our concept of adverse possession
of the intellectual property space comes into play. Just as real property
owners have a right to exclude others from their property, intellectual
property rights should also confer a right to exclude. Open and notorious occupation of real property, however, can confer ownership to the
trespasser after a given period of time. As discussed earlier, we propose that the same principle be extended to the intellectual property
348. Id. at 822.
349. Id.
350. See Apple Kills Clone Market, WIRED NEWS, Sept. 2, 1997, available at
http://www.wired.com/news/business/0,1367,6548,00.html (last visited Mar. 13, 2003).
351. See supra notes 92–94 and accompanying text.
352. The issue of sunk costs could then be dealt with contractually.
374
Harvard Journal of Law & Technology
[Vol. 16
space. Just as the original property owner loses the right to eject those
who successfully adversely possess the real property in question, if a
firm allows an ancillary market to develop, it should not be able to
later extinguish that market by changing its policy of intellectual
property right enforcement. This adverse possession concept is consistent with Kodak I, in that even if the firm does not have market power
in the primary product market, it may have market power in the secondary market if there are high switching costs and it permits a secondary market to develop.353
The prohibition against making an important change in a pattern
of distribution that had originated in a competitive market and had
persisted for a given period of time, as in Aspen Ski, only applies retroactively. The firm would retain the right to exclude the development
of an ancillary market for future models of existing products or new
products. Just as Ski Co. could have opened a new mountain without
including it in the All-Aspen ski pass program, the firm could sell
future primary products and maintain exclusive control of any potential ancillary market for those products.
Under this proposal, customers who purchased products when the
possibility of an ancillary market existed would thus not be disadvantaged, whereas customers who purchased new products after the firm
announced a policy of not selling replacement parts to ISOs would do
so with the full knowledge of the unavailability of the ancillary market for their specific purchase. These new customers would be able to
factor the lack of an ancillary market into their purchase decision.
Despite its environmental unpalatability, a good analogy is the
creation of new coastal land through the process of dredging, such as
occurred with Boston’s Back Bay. Once the land was created, it theoretically would have been possible for someone, through open and
notorious occupation of the landfill areas, to obtain rights in that land
through adverse possession. That right of occupancy would not extend, however, to lands subsequently created unless there was similar
open and notorious occupation of that new land. If the property creator excluded the prior adverse possessor from the new land, no adverse possession of that new land would take place.
Continuing with our expanded real property analogy, just as there
may be public policy reasons to grant easements or rights-of-way
across this newly created coastal landfill property for development
purposes, it may be necessary to grant easements in the intellectual
property space.354 Although this situation seems more likely in the
353. Note that in this third scenario, if access to protected intellectual assets is initially
required for such ancillary markets to arise, the cooperation of the firm is still required. If
the ancillary market arises due to infringement, our proposal would not apply.
354. See infra Part V.A. for discussion of the applicability of the easements analogy.
No. 2]
Adverse Possession for Intellectual Property
375
copyright context given the fair use doctrine,355 and less likely in the
patent context given the Federal Circuit’s narrow interpretation of
experimental use,356 our expanded notion of essential facilities discussed in detail in the next section could be considered an easement
by necessity.357
An open question remains as to whether any refusal to deal could
include a refusal to sell unpatented parts. If the firm in this instance
does not have market power, such a refusal would not seem to run
afoul of antitrust principles. If the firm had allowed an ancillary market to arise, however, it is possible that it could be found to have market power, as in Kodak I. In that event, the issue of refusing to deal in
unpatented parts might be more problematic, and we would propose a
Rule of Reason analysis.
3. Firms with Market Power
If a firm has market power in any market, the same three options
would still be available for the introduction of a new product. The
firm could act to prevent any ancillary market from arising by refusing to sell or license products embodying its intellectual property, or it
could grant short term contracts to provide service. If the firm allowed
the ancillary market to develop, however, it would be subject to the
same adverse possession rule as the firm without market power.
Because the firm would also have market power in the primary
market, its ability to exclude other firms from the ancillary market ex
ante is subject to an additional constraint that is a variation of the Essential Facilities Doctrine. Historically, an essential facility has been
found only in instances involving direct competitors. In fact, one of
the core requirements for a facility to be deemed essential is “a [direct
or horizontal] competitor’s inability practically or reasonably to duplicate the essential facility.”358 In this instance we propose that the
doctrine be extended to downstream competitors359 but only in industries that exhibit network effects and only for new generations of
technology where the prior technology could have been characterized
as an essential facility. In this instance, the Rule of Reason would ap355. See supra note 20 and accompanying text.
356. See supra note 28 and accompanying text.
357. See infra notes 407–10.
358. See MCI Communications Corp. v. Am. Tel. & Tel. Co., 708 F.2d 1132 (7th Cir.
1983).
359. An example of a situation involving downstream competitors would be Microsoft,
as developer of an operating system, relative to an applications developer that develops a
competing product to one also offered by Microsoft (such as a word processor or spreadsheet). Relative to Microsoft, both Corel and Lotus would be examples of downstream
competitors in the applications market that rely on Microsoft’s operating system as an essential facility. See infra Part IV.B.3 for a detailed examination of the application of essential facilities to downstream competitors in industries that exhibit network effects.
376
Harvard Journal of Law & Technology
[Vol. 16
ply. Part of the rationale for this extension is to maintain and enhance
the incentives for interoperability.
Our extension of Essential Facilities Doctrine is based in part on a
civil antitrust case in which one of the authors testified as an expert
witness against Microsoft.360 In that case, a small software manufacturer argued that Microsoft’s Windows 95 was an essential facility.
The U.S. District Court for the Southern District of Texas stated that
an operating system such as Windows “could be an essential facility
for application software, such as a word processing program,”361 but
not for a software product “whose sole purpose is to improve on . . .
the facility at issue.”362 Therefore, the application of the Essential Facilities Doctrine turned on whether the software product ran on top of
the operating system or inside of it.363 If the latter, Essential Facilities
Doctrine would have been inapplicable.364 If the former, then Essential Facilities Doctrine might have applied.365
Although not addressed by the court, the logical conclusion of
this opinion suggests the extension of the Essential Facilities Doctrine
to downstream competitors because Microsoft was already engaged in
the development and marketing of word processing applications. Nor
should our proposed extension of essential facilities be limited to Microsoft operating systems.366
The question of whether the firm could refuse to sell unpatented
spare parts still remains, but the answer is likely to be different for a
firm with market power. A refusal to sell unpatented parts could be an
instance of tying, but it could alternatively be a different unilateral
refusal to deal. If the only use for the non-patented parts were to service the primary product, which would entail using aspects of the
product protected by the intellectual property laws, then would the
firm be entitled to withhold the unpatented parts? Nothing would preclude a separate manufacturer from making the unpatented parts, but
both it and the ISO that installs them might be infringers (the manufacturer being a contributory infringer), depending on whether the
firm sells or licenses the primary product. Thus, we would again propose a Rule of Reason analysis for any refusal to deal in unpatented
parts.
360. David L. Aldridge Co. v. Microsoft Corp., 995 F. Supp. 728 (S.D. Tex. 1998) (Mr.
Clarkson testified on behalf of the plaintiff.).
361. Id. at 754.
362. Id.
363. Id. at 753–54.
364. Id. at 756.
365. Id.
366. For an example outside the Microsoft context, see infra Part IV.B.3 regarding a
variation of the Intergraph case.
No. 2]
Adverse Possession for Intellectual Property
377
B. Application to Unilateral Refusal to Deal
Having developed our proposal establishing the constraints on a
firm’s ability to refuse to license or sell items embodying its intellectual property, we now proceed to evaluate several relevant fact patterns to show how our proposal would be applied.
1. Large Fixed Assets (the Kodak and Xerox scenarios)
Under our proposal, a firm that offers a large fixed asset product
to the market has three choices:
• Announce a policy that it will not sell replacement
parts to anyone, so that consumers buying the product will not expect ISO service to be available.
• Enter into exclusive ISO contracts of relatively
short duration that will be put up to bid subsequently
so that there is competition in the contract (a situation akin to a lease of real property).
• Sell replacement parts to anyone.
Kodak chose the third option, but then later changed to the first
option. Under our proposal, the ruling in Kodak I would remain the
same. While the outcome of Kodak II would also be the same, the
determining factor would be entirely different. Whereas the Ninth
Circuit focused on the subjective intent of Kodak’s management, our
proposal eschews the examination of intent altogether. Unlike the intent examination, which would almost always require that the matter
go to a jury, the adverse possession determination could be made by a
judge as a matter of law pursuant to a motion for summary judgment.
If Kodak introduced a new type of copier, given that Kodak does
not have market power in the copier market, it would be permitted to
sell its machines under a policy announced at the time of sale that
there will be no replacement parts sold to ISOs. Even if Kodak initially were willing to sell replacement parts to ISOs, Kodak would be
permitted to change that policy for new models going forward but not
retroactively for existing models. In other words, if Kodak permitted
the ISOs to occupy part of the space that Kodak could have occupied
initially, then Kodak could not now extinguish that market. If this
were not the case, Kodak could impose costs on consumers not contemplated at the time of purchase. This restriction means that when
consumers are deciding which copier to buy, they need to take into
account the effect of Kodak’s service contract fees when deciding
378
Harvard Journal of Law & Technology
[Vol. 16
whether to purchase the Kodak copier or a copier produced by another
manufacturer.
Our analysis of the facts of the Xerox case produces an outcome
similar, but not identical, to that of the Federal Circuit’s decision. Our
proposal would allow Xerox to continue its policy of refusing to sell
parts to the ISOs, but it would differ significantly from the Federal
Circuit’s decision in its treatment of Xerox’s retroactive application of
the policy. Xerox would be able to refuse to sell parts to ISOs for service on new models sold after the policy change, but it could not refuse to sell parts for copiers sold during the time of the previous
policy. While the ISOs would be foreclosed from servicing a part of
the future market, they would still be able to service existing customers; thus, those customers would also be protected. In either instance,
the copier industry clearly does not exhibit network effects, so our
modified Essential Facilities Doctrine would not apply.
2. Creation of a New Product Market: Segway
The Segway Human Transporter (“HT”) is one of inventor Dean
Kamen’s latest innovations. The Segway HT is a self-balancing, electric powered personal transportation machine that emulates human
balance, which is controlled by subtle shifts in body weight.367 Prior
to its unveiling last December, the product was codenamed “Ginger”
in order to keep the identity of the device secret.368 Since then, the
“Segway HT has gained widespread interest from consumers and
businesses around the globe.”369
The Segway HT is protected by a number of patents,370 and the
device is arguably sufficiently unique as to define a market in and of
itself.371 In that market, Segway LLC clearly has market power, but its
initial three options would remain the same as for a firm without market power: no ISOs, contracted ISOs, or selling parts to anyone.
Currently it appears that Segway has chosen option number one,
although it is offering to provide training for service technicians.372
Although Segway clearly has market power in the market for selfbalancing, electric powered personal transportation machines, such a
367. Press Release, Segway LLC, $500,000 Lemelson-MIT Prize Awarded to Dean
Kamen (Apr. 23, 2002), at http://www.segway.com/aboutus/press_releases/pr_042302b.
html.
368. Id.
369. Id.
370. A quick LEXIS search identified at least seventeen patents registered to Dean
Kamen related to personal balancing vehicles.
371. It could also be viewed as a substitute for bicycles or motorized scooters, but for our
purposes here, we assume the market for self-balancing transporters is sufficiently distinct
so as to be a separate market.
372. See Segway Support at http://www.segway.com/support/ (last visited Feb. 19,
2003).
No. 2]
Adverse Possession for Intellectual Property
379
market does not exhibit network effects, and thus our modification to
the Essential Facilities Doctrine would not apply.
3. Operating Systems and APIs (Microsoft)
Copiers and other capital equipment have long service lives and
are often replaced when they wear out. Operating systems, however,
never wear out and are instead replaced when a new version is released, often with a relatively brief time between releases. Switching
costs for copiers are based primarily on the cost of the new copier
with the labor costs being a small fraction of total costs. Conversely,
the cost of a software upgrade is usually fairly small; the bulk of any
switching costs are a function of the labor cost associated with both
installing the new software and training the user how to operate it.
The switching costs for operating systems are increased dramatically, however, if all of the applications software must be replaced. To
reduce those costs, most software development in the last decade or so
has been done using connections to the operating system called Application Programming Interfaces (“APIs”). As the D.C. Circuit
pointed out:
Operating systems . . . function as platforms for
software applications . . . by “exposing” — i.e., making available to software developers — routines or
protocols that perform certain widely-used functions
[called APIs]. For example, Windows contains an
API that enables users to draw a box on the screen.
Software developers wishing to include that function
in an application need not duplicate it in their own
code. Instead, they can “call” — i.e., use — the Windows API. Windows contains thousands of APIs,
controlling everything from data storage to font display.
Every operating system has different APIs. Accordingly, a developer who writes an application for one
operating system and wishes to sell the application to
users of another must modify, or “port,” the application to the second operating system. This process is
both time consuming and expensive.373
A potential problem arises when the operating system programmers work for a different company from some, but not all, of the ap373. Microsoft 2001, 253 F.3d 34, 53 (D.C. Cir. 2001).
380
Harvard Journal of Law & Technology
[Vol. 16
plications programmers. If the operating system company provides its
own applications programmers earlier access to the APIs than it provides to applications programmers at other companies, then those
other companies would be at a tremendous competitive disadvantage.
The Justice Department alleged a variation of this fact pattern in
its first suit against Microsoft where Microsoft purportedly conditioned access to APIs on agreements by software manufacturers not to
work with competing technologies.374 However, the allegation was
neither part of the first consent decree approved in 1995 nor raised as
an issue in the subsequent litigation. It seems that Judge Jackson’s
remedy of breaking up Microsoft into an operating systems company
and an applications company evidenced his concern that Microsoft
was leveraging its dominance in the OS market to obtain monopoly
power in the applications market. The requirement imposed by the
Final Judgment in United States v. Microsoft Corp.,375 however, that
Microsoft provide rival applications firms earlier access to its Windows APIs appears to eliminate at least some of the advantages enjoyed by the Microsoft applications group.
Operating systems clearly demonstrate network effects, given that
the more users there are for a particular operating system, the more
applications are likely to be written to run on top of that operating
system.376 In his findings of fact, Judge Jackson described the situation as follows:
Consumer demand for Windows enjoys positive
network effects. A positive network effect is a phenomenon by which the attractiveness of a product increases with the number of people using it. The fact
that there is a multitude of people using Windows
makes the product more attractive to consumers. The
large installed base attracts corporate customers who
want to use an operating system that new employees
are already likely to know how to use, and it attracts
academic consumers who want to use software that
374. See Complaint of United States at 6–7, Microsoft 2001, 253 F.3d 34 (D.C. Cir.
2001), http://www.usdoj.gov/atr/cases/f0000/0046.htm (last visited Feb. 19, 2003).
375. 2002 U.S. Dist. LEXIS 22864 (D.D.C. 2002).
376. See Abramson, supra note 188 at 173; see also, e.g., Richard J. Gilbert & Michael
L. Katz, An Economist’s Guide to U.S. v. Microsoft, UC Berkeley Working Paper (2001), at
http://iber.berkeley.edu/wps/econ/E01-300.pdf; Mark A. Lemley & David McGowan, Legal
Implications of Network Economic Effects, 86 CAL. L. REV. 479 (1998); Carl Shapiro, Antitrust in Network Industries, address before the American Law Institute and American Bar
Association, delivered Jan. 25, 1996 (transcript available at http://www.usdoj.gov/atr/
public/speeches/shapir.mar); Michael L. Katz & Carl Shapiro, Systems Competition and
Network Effects, 8 J. ECON. PERSP. 93 (1994).
No. 2]
Adverse Possession for Intellectual Property
381
will allow them to share files easily with colleagues
at other institutions.377
The D.C. Circuit referred to this phenomenon as a “chicken-andegg situation[, which] ensures that applications will continue to be
written for the already dominant Windows [platform], which in turn
ensures that consumers will continue to prefer it over other operating
systems.”378
Therefore, if an operating system manufacturer developed an upgrade to its operating system and refused to give outside developers of
applications software the same access to the new APIs as it gave to its
internal applications software developers, application of our framework would likely result in the operating system manufacturer being
compelled to grant access to the APIs on a prompt, reasonable, and
non-discriminatory basis.379
Whereas we might not be able to achieve such a result based on
monopoly leveraging,380 application of essential facilities in the case
of a product exhibiting network effects produces the appropriate result. Our proposal would thus lead to the same result as the Final
Judgment in United States v. Microsoft Corp.381
While Microsoft is the most obvious scenario for the application
of our proposed extension of essential facilities, it is by no means the
only one. If Intel had manufactured graphical workstations at the time
of the Intergraph case, our proposal would have suggested382 that access to the “special strategic customer benefits”383 would have in fact
been an essential facility.
4. European Application
The issues addressed in this Article are not unique to the United
States, and the European Union has developed its own jurisprudence
377. United States v. Microsoft Corp., 84 F. Supp. 2d 9, 20 (D.D.C. 1999).
378. Microsoft 2001, 253 F.3d at 55 (internal quotation marks omitted).
379. The question of what is a reasonable royalty is non-trivial. The Federal Circuit has
stated that the determination of “a fair and reasonable royalty is often . . . a difficult judicial
chore, seeming often to involve more the talents of a conjurer than those of a judge.” Fromson v. Western Litho Plate & Supply Co., 853 F.2d 1568, 1574 (Fed. Cir. 1988). As Judge
Learned Hand put it, “the whole notion of a reasonable royalty is a device in aid of justice,
by which that which is really incalculable shall be approximated.” Cincinnati Car Co. v.
New York Rapid Transit Corp., 66 F.2d 592, 595 (2d Cir. 1933).
380. See supra Part II.E.
381. 2002 U.S. Dist. LEXIS 22864 (D.D.C. 2002). The 2002 Final Judgment does not,
however, rely on essential facilities as a rationale.
382. The district court found an essential facility in Intergraph Corp. v. Intel Corp., 3 F.
Supp. 2d 1255, 1278 (N.D. Ala. 1999), but incorrectly so, as there was no competition between Intel and Intergraph in any market.
383. See discussion of the Intergraph case, supra Part III.E.
382
Harvard Journal of Law & Technology
[Vol. 16
regarding such matters.384 Although our framework seems to be more
specific than the European approach, the two can be reconciled. In the
European Union, the European Commission held that it was unacceptable for IBM to tie sales of its CPU to sales of its memory and
service.385 In contrast, the Commission said that it was acceptable for
Kyocera, a Japanese computer printer company, to offer discounted
bundles of Kyocera-made printer toner cartridges and other Kyoceramade printer hardware even though the bundle might otherwise have
been characterized as an impermissible economic tie.386 Pelikan, a
German manufacturer of toner cartridges for use with Kyocera printers, had challenged both the bundles and Kyocera’s policy of not honoring warranty claims for printers damaged by non-Kyocera toner
cartridges as an abuse of Kyocera's dominant position.387 The Commission found: (1) that Kyocera did not have a dominant position in
the market for toner cartridges despite its large share of the toner market; (2) that there was fierce competition in the primary market for
printers that restrained Kyocera’s behavior in the secondary market
for toner; (3) that purchasers of printers were well informed about the
price charged for toner and appeared to take this into account when
selecting a printer; and (4) that the cost of toner was a significant part
of the cost of the entire product.388 As a result, if Kyocera had attempted to raise the price of its toner cartridges, then the low switching costs would lead customers to junk the printer and buy another
brand for which cheaper toner cartridges were available.389 This contrasts sharply with the situation in Kodak I when Kodak changed its
384. See TREATY ESTABLISHING THE EUROPEAN COMMUNITY, Nov. 10, 1997, O.J. (C
340) 3 (1997) [hereinafter EC Treaty]. Article 81 of the EC Treaty (originally Article 85,
but subsequently renumbered) prohibits undertakings from entering into agreements that
have as their object or effect the restriction, prevention, or distortion of competition within
the Common Market. Article 82 (originally Article 86) prohibits the abuse by undertakings
of a dominant position in a particular market to the detriment of trade between Member
States. A dominant position “has always been defined by the Commission as the ability to
act to an appreciable extent independently of competitors and consumers. Therefore, an indepth fact-finding exercise and analysis on a case-by-case basis are required.” Commission
of the European Communities, 25th Report on Competition Policy 41 (1995), available at
http://europa.eu.int/comm/competition/publications/ra9501en_en.pdf (last visited Feb. 19,
2003). Generally, “examination of the characteristics of a sector of activity in order to decide whether or not it constitutes a market within the meaning of Article 86 of the Treaty
cannot be carried out in the abstract, but on the contrary must be established on a case-bycase basis with the aim of analysing the behaviour of the undertaking in question in the light
of competition law.” Commission Decision Relating to a Proceeding Pursuant to Articles 85
and 86 of the European Economic Community Treaty 94/210, art. 116, 1994 OL (L 104) 34.
385. See Commission of the European Communities, 14th Report on Competition Policy,
78 (1984).
386. See Commission of the European Communities, 25th Report on Competition Policy
41 (1995) (1995), available at http://europa.eu.int/comm/competition/publications/
ra9501en_en.pdf.
387. Id.
388. Id.
389. Id.
No. 2]
Adverse Possession for Intellectual Property
383
policy after customers were locked into a high cost capital asset with
high switching costs.
Currently before the European Court of the First Instance is a case
somewhat similar to the facts in Data General.390 In IMS Health Inc.
v. Commission of the European Communities, IMS Health developed
a sophisticated method of tracking pharmaceutical sales involving a
new data structure.391 In November and December of 2000, IMS successfully sued two other competitors for copyright infringement before the Landgericht Frankfurt and obtained an injunction prohibiting
them from using the new data structure.392 The Oberlandesgericht
Frankfurt subsequently overturned the injunction.393 The Commission
of the European Communities then imposed a compulsory licensing
requirement,394 concluding both that IMS’s data structure was an essential facility395 and that IMS’s “refusal to license the use of its copyright . . . constitutes an abuse of the dominant position it enjoys in the
relevant market.”396 IMS then appealed to the Court of the First Instance, which suspended the injunction until the main action is decided.397
Under our test, for IMS to be found liable for refusing to license
new technology in the pending case, IMS would have to be in a position of market power and its copyrighted data structure would have to
be both an essential facility and a network effects type technology.
While IMS seems to be in a position of market power,398 given that
the two other competitors “already have a significant number of customers and have both developed alternative (and allegedly noninfringing) [data] structures that they can continue to use and improve,”399 it would appear unlikely that the new data structure would
be found to be both an essential facility and to exhibit significant network effects.
C. Application to Unilateral but Conditional Refusal to Deal
At first glance it would seem logical that if an intellectual property right holder has the right to unilaterally refuse to deal with someone, that right would include the right to conditionally refuse to deal
with that same someone, particularly given that conditional dealing
390. Data Gen. Corp v. Grumman Corp., 761 F. Supp. 185 (D. Mass. 1991). For our discussion of the case, see supra Part III.E.3.
391. Case T-184/01, 1 E.C.R. 3193 (2001).
392. Id. at 6–8.
393. Id. at 10.
394. Id. at 12–18.
395. Id. at 11.
396. Id. at 47.
397. Id. at 104.
398. Id. at 14.
399. Id. at 92.
384
Harvard Journal of Law & Technology
[Vol. 16
might be more accommodating. As economist Carl Shapiro has
pointed out, however, it is possible to “create incentives through conditional refusals that you can’t create through unconditional refusals.”400
While our framework was designed to answer questions regarding
unilateral (or unconditional) refusals to deal, it may still provide some
insight for conditional refusals to deal when we reexamine the various
scenarios examined above.
1. Large Fixed Assets
The possibility of adverse possession of certain elements of the
intellectual property grant would still apply in this instance, and so the
analysis would turn on whether incentive modifications resulting from
the conditional refusal to deal are such that they result in an impermissible recapture of the intellectual property right space that was
previously lost to adverse possession. If the firm has market power, its
attempt to recapture that space through the imposition of conditions
are more likely to be categorized as either tying or monopoly leveraging.
2. Segway
The analysis for this scenario remains the same, as long as Segway continues to foreclose the possibility of ISOs. As with the prior
example, however, if Segway allows a secondary market to develop, it
runs the risk of liability for either tying or monopoly leveraging if it
attempts to recapture that secondary market through conditional refusals to deal.
3. Operating Systems
It is hard to discuss this scenario without being heavily influenced
by the D.C. Circuit’s decision in Microsoft 2001.401 We would argue
that in Microsoft 2001, the court misinterpreted the Xerox opinion.
Although the court correctly cited Xerox for its statement “intellectual
property rights do not confer a privilege to violate the antitrust
laws,”402 it seemed to ignore the very next sentence in the Xerox opin-
400. See Department of Justice Antitrust Division and Federal Trade Commission, Hearing on Competition and Intellectual Property Law and Policy in the Knowledge Based
Economy, “Strategic Use of Licensing: Is There Cause for Concern about Unilateral Refusals to Deal?” 156, May 1, 2002, available at http://www.ftc.gov/opp/intellect/
020501xscript.pdf.
401. Microsoft 2001, 253 F.3d 34 (D.C. Cir. 2001).
402. Id. at 63 (quoting Xerox, 203 F.3d 1322, 1325 (Fed. Cir. 2000)) .
No. 2]
Adverse Possession for Intellectual Property
385
ion: “it is also correct that the antitrust laws do not negate the patentee’s right to exclude others from patent[ed] property.”403
Many of the allegations against Microsoft that could be characterized as conditional refusals to deal, however, were “garden variety”
tying and monopoly leveraging in areas in which Microsoft was not
asserting intellectual property protection. Given that our framework
only attempts to address instances where intellectual property rights
form the basis for refusal to deal, either unconditional or conditional,
we have not addressed those other garden variety areas in this Article.
To the extent that Microsoft sought to justify limiting its OEMs’
ability to change the user interface or its policy of selling Windows at
a lower price to OEMs who agreed not to install competing software,
the D.C. Circuit correctly characterized those activities as willful attempts to maintain Microsoft’s monopoly on Intel-compatible PC operating systems. The primary evil the Final Judgment sought to
address by requiring uniform licenses was Microsoft’s attempt to deter OEMs from dealing with Microsoft’s competitors. Such conduct
would appear to be impermissible even in the Federal Circuit given
the Federal Circuit’s reasoning in Xerox that a patent infringement
suit or counter-claim is not immune from antitrust scrutiny when it is
a mere sham to disguise an attempt to interfere directly with the business relationship of a competitor.
V. POSSIBLE CRITIQUES
Critiques of our proposal are likely to come from two different
perspectives. The first concerns the appropriateness of the adverse
possession analogy itself. The second involves the practicability of
our proposal. We discuss each in more detail in the following two
sections.
A. Real Property Critique — Why Not Easements for Intellectual
Property?
Given that we are proposing the application of adverse possession
to intellectual property, one might ask why is not easement a better
analogy for what we are trying to accomplish? To answer this question, we need to look at the different types of easements and the way
such easements are created.
With respect to use, easements can be classified as either positive
or negative.404 Positive, or affirmative, easements give the holder of
403. Xerox, 203 F.3d at 1325 (quoting Intergraph Corp v. Intel Corp, 195 F.3d 1346,
1362 (Fed. Cir. 1999)).
404. See HERBERT HOVENKAMP & SHELDON F. KURTZ, THE LAW OF PROPERTY 320
(2001).
386
Harvard Journal of Law & Technology
[Vol. 16
the easement a right to do something on someone else’s land,405 thus
requiring the original owner to share access to the land to some degree. Conversely, negative easements, or restrictive covenants, give
the holder of the easement the right to keep a landowner from doing
something on her own land.406 Like adverse possession, negative
easements can be used to preclude the original owner from exercising
the full complement of property rights in the original space. Whether
positive or negative, easements can be created by implication, prescription, necessity, estoppel, express provision, or eminent domain.407 We discuss the first three types as they might apply to
intellectual property.408
Implied easements are generally created when an express agreement between the parties is either silent or ambiguous as to whether
the grantor intended to create an easement, or when a formal easement
effectuating the intent of the parties was omitted by mistake.409 Such
easements are usually associated with sales, transfers, or conveyances
of land.410 An example would be where the owner of both parcel A
and B has historically crossed parcel A to reach the main highway
from parcel B. That owner subsequently sells parcel B to another
owner but neglects to expressly create an easement to allow the new
owner to access the highway. The new owner of parcel B would be
entitled to an easement across parcel A along the same path as was
previously used by the original owner to reach the road.411 Such an
405. See SINGER, supra note 315, at 171; see also HOVENKAMP & KURTZ, supra note
405, at 320 (affirmative easements “entitle the easement owner to do affirmative acts on the
land in the possession of another”).
406. See SINGER, supra note 315, at 171; see also HOVENKAMP & KURTZ, supra note
404, at 320 (negative easements “take from the owner of the servient estate the right to do
some things which, were it not for the easement, she would have a right to do on her own
land”).
407. See HOVENKAMP & KURTZ, supra note 404, at 321. Hovenkamp and Kurtz classify
easements by implication and by necessity together as easements by implication. Id. at 322–
23. Professor Singer suggests a different classification scheme whereby easements by estoppel, easements by necessity, and easements implied by prior use are collectively referred
to as easements by implication, which he also calls quasi-easements. See SINGER, supra
note 315, at 176–89.
408. Easements by estoppel usually arise in the context of a real property license. See
SINGER, supra note 315, at 177. Express easements in the intellectual property context
would also be in the form of a license, and intellectual property licenses already have a
separate place in the antitrust enforcement regime. See, e.g., United States Dep’t of Justice
& Fed. Trade Comm’n, Antitrust Guidelines for the Licensing of Intellectual Property
(1995), available at http://www.usdoj.gov/atr/public/guidelines/ipguide.htm. Easements by
eminent domain seem inapplicable to the scenarios at issue. For a discussion of such easements in the intellectual property context, however, see generally Thomas F. Cotter, Do
Federal Uses of Intellectual Property Implicate the Fifth Amendment? 50 FLA. L. REV. 529
(1998).
409. See SINGER, supra note 315, at 176, 185.
410. Id.; see also HOVENKAMP & KURTZ, supra note 404, at 321 (easements by implication are “created by law in connection with a conveyance of part of a tract of land”).
411. See, e.g., Granite Props. Ltd. P’ship v. Manns, 512 N.E. 2d 1230, 1236 (Ill. 1987)
(Easements implied from prior use arise when the owner of two or more adjoining parcels,
No. 2]
Adverse Possession for Intellectual Property
387
easement would be implied from prior use, and such easements are
generally limited to situations where there has been a transfer or conveyance of property.412
In the intellectual property situations that we have examined, this
construction does not seem to apply. The original owner is not conveying ownership of the ancillary market space, but is instead permissively allowing occupancy of that space by firms such as the ISOs.
Although it is true that patented spare parts are being sold to the ISOs,
such sales do not convey the ancillary market space for servicing the
equipment itself. If the original owner sold its service business to an
ISO, the ISO might reasonably expect a continued supply of patented
spare parts. That fact pattern, however, would be an entirely different
situation from the scenarios that are at issue here. Thus, an analogy of
easements implied from prior use seems less appropriate than our adverse possession analogy.
The second type of easement is an easement by prescription. An
easement by prescription is a positive right relative to another owner’s
land that is obtained through the same principles as adverse possession.413 Unlike the doctrine of adverse possession, however, whereby
the original owner may be precluded from certain activities that were
allowable before the adverse possession took place, negative easements cannot be acquired by prescription.414 Accordingly, for easements by prescription, the original owner is not excluded from fully
participating415 in the space covered by the easement.416
Under Kodak I,417 if the original owner of the ancillary market
has market power in the primary market, then the original owner may
be precluded from engaging in certain types of anticompetitive conduct, such as tying and exclusive dealing.418 An analogy of a prescriptive easement in the ancillary market space could not preclude the
original owner from engaging in tying to exclude competition, howwhere one parcel derives from the other a benefit or advantage of an apparent, continuous,
and permanent nature, transfers or conveys one of the parcels without mention being made
of these incidental uses. “In the absence of an expressed agreement to the contrary, the
conveyance or transfer imparts a grant of property with all the benefits and burdens which
existed at the time of the conveyance of the transfer, even though such grant is not reserved
or specified in the deed.”).
412. See HOVENKAMP & KURTZ, supra note 404, at 322 (Easements by implication can
only be established if “at the time of the conveyance one part of the land is being used for
the benefit of the other part.” Absent a conveyance, there can be no implied easement.).
413. See SINGER, supra note 315, at 190; see also HOVENKAMP & KURTZ, supra note
404, at 322 (“Easement by prescription arises by adverse use of the servient estate by the
dominant tenant for the period of the statute of limitation used for adverse possession.”).
414. SINGER, supra note 315, at 197.
415. Minus the right to exclude the holder of the easement from the relevant space.
416. This limitation would also exist for both easements implied from prior use and easements by necessity.
417. 504 U.S. 451, (1992).
418. See supra Part III.A.
388
Harvard Journal of Law & Technology
[Vol. 16
ever, because such a preclusion of the original owner would constitute
a negative prescriptive easement, which would be invalid. Unlike an
easement, the adverse possession analogy would allow such activity
to be prevented.419
For example, an original owner such as Segway can exclude all
other sources of service for its devices because there is no ancillary
market for service that is separate and distinct from the market for
Segway devices. However, if Segway allows such an ancillary market
to arise, subsequent attempts to force customers to get service only
from Segway would be a case of impermissible tying under our adverse possession analogy, but not under an easement analogy.420
Although the adverse possession analogy is preferable to an
easement analogy in the intellectual property space for existing products, our extension of the Essential Facilities Doctrine to the intellectual property space for new products may be analogous to another
type of implied easement, an easement by necessity. Easements by
necessity generally arise out of the need to prevent landlocked parcels
of real property.421 Some scholars have suggested that if application of
the Essential Facilities Doctrine required the license of an intellectual
property right,422 then the result would be an “easement across the
monopolist’s intellectual property.”423 In the applications software
context, for example, the software developers own their own intellectual property space, but if they are denied access to the APIs for the
next generation of the relevant operating system, they could be
viewed as being landlocked in their intellectual property space with
no way to reach the customers who upgrade to the new operating system. We believe that an application of the Essential Facilities Doctrine
that allowed the applications programmers to access the APIs could
419. Some of the potentially difficult issues regarding the specific elements of adverse
possession, such as permissive use, or what constitutes adverse or hostile occupation, are the
same for prescriptive easements. Those issues have already been addressed in supra Part
IV.A.1.
420. Under either analogy, however, a reseller of used Segways could tie sales and service without restriction, as long as that reseller is not in a position of market power.
421. See SINGER, supra note 315, at 187. Singer also points out that prior use is not required to obtain an easement by necessity. Id. Note that Hovenkamp and Kurtz limit easements by necessity to those situations “when property becomes landlocked by virtue of a
conveyance. [In that situation,] the landowner who can show strict necessity is entitled to a
right of way across the severing line that made the property landlocked.” HOVENKAMP &
KURTZ, supra note 404, at 323.
422. See Maureen A. O'Rourke, Striking a Delicate Balance: Intellectual Property, Antitrust, Contract, and Standardization in the Computer Industry, 12 HARV. J.L. & TECH. 1
(1998).
423. Teague I. Donahey, Terminal Railroad Revisited: Using the Essential Facilities
Doctrine to Ensure Accessibility to Internet Software Standards, 25 AM. INTELL. PROP.
LAW ASS’N Q.J. 277, 313 (1997).
No. 2]
Adverse Possession for Intellectual Property
389
appropriately be characterized as an easement,424 as could the fair use
doctrine applicable to copyrights.425
B. Intellectual Property and Antitrust Critiques
Although we feel that our adverse possession analogy is preferable to an easement analogy, several potential critiques still exist relative to the applicability of our proposal. One such possible critique of
our extension of essential facilities to downstream competitors is that
such an application would necessitate a reversal of Berkey Photo.426
Given that we constrain our extension to existing markets that exhibit
network effects, however, the introduction of a new type of camera/film combination would not qualify for such an extension of essential facilities.
One might also point out that the Federal Circuit has been reluctant to extend the Essential Facilities Doctrine to downstream competitors, noting in Intergraph that
no court has taken it beyond the situation of competition with the controller of the facility, whether the
competition is in the field of the facility itself or in a
vertically related market that is controlled by the facility. That is, there must be a market in which plaintiff and defendant compete . . . . Absent such a
relevant market and competitive relationship, the essential facility theory does not support a Sherman
Act violation.427
At the time of the Federal Circuit opinion, Intel and Intergraph
were not competitors,428 and thus the district court’s ruling that “Intel's superior microprocessor product and Intergraph’s dependency
thereon converted Intel’s special customer benefits into an ‘essential
facility’ under the Sherman Act [could] not be sustained.”429 Unlike
424. Reverse engineering, however, would be closer to the fair use situation in the Sega
v. Accolade case, discussed supra in Part III.E.3.
425. Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263 (2d Cir. 1979), cert. denied, 444 U.S. 1093 (1980). See supra text accompanying notes 355–356.
426. See supra Part II.E for discussion.
427. Intergraph Corp. v. Intel Corp., 195 F.3d 1346, 1357 (Fed. Cir. 1999).
428. Hovenkamp, Janis, and Lemley suggest, however, that the Federal Circuit was “arguably incorrect to conclude that the parties were not in the same market, [given that] Intergraph’s intellectual property rights, the assertion of which triggered the dispute, were in the
same technology market as Intel’s primary line of business (microprocessors)” (emphasis
added). The authors do point out that the Federal Circuit’s “failure to recognize this does not
affect the essential facilities analysis, however, because any such competition would exist in
the market for the essential facility itself, not the downstream market Intel was allegedly
trying to control.” IP AND ANTITRUST, supra note 1, § 13.3c2 n.23.
429. Intergraph. at 1358; see also supra Part III.E (discussing Intergraph).
390
Harvard Journal of Law & Technology
[Vol. 16
the situation in Intergraph, however, we limit our extension of essential facilities to the case where the holder of the essential facility does
compete in the downstream market.430
The district court in Data General also cautioned against an expansive application of essential facilities:
The “bottleneck” of [Data General’s] superior
knowledge in the design of [its own] computers is
insufficient to invoke the essential facilities doctrine;
a better mousetrap is not necessarily an essential facility. The Sherman Act has not been interpreted to
require manufacturers to abandon their advantage in
creating accessories to their systems. If manufacturers of complex and innovative systems were required
to share with competitors the development of accessories, because they had a possibly absolute advantage through producing the system, the incentives of
copyright and patent laws would be severely undermined. Not only would the manufacturer, who is in
the best position to create these accessories, have
less incentive to do so, but also the impetus for competitors to reverse engineer and produce competing
solutions would be reduced.431
We agree wholeheartedly and thus limit our extension of the Essential Facilities Doctrine to situations where the essential facility
owner competes directly in the downstream markets, and only then in
industries exhibiting network effects. We would not characterize copier spare parts or software diagnostics as essential facilities, but we
would characterize operating system APIs necessary for developing
software applications as essential facilities. Thus the application of
our proposed extension is quite narrow.432
One could also argue that our adverse possession argument could
be applied to the problem of patent nonuse or patent suppression433
and thus be used to deprive firms of strategic options regarding their
intellectual property portfolios. Our proposal is limited, however, to
arenas where the patents in question are already in use by either their
owners or their licensees and the issue is whether other firms should
also be allowed to make use of the patented technology. The idea of
430. As would be the case with Microsoft and other desktop applications developers.
431. Data Gen. Corp v. Grumman Corp., 761 F. Supp. 185, 192 (D. Mass. 1991).
432. Although not so narrow as to be limited only to the facts of the Microsoft litigation.
433. For an excellent discussion of this subject, see Saunders, supra note 238.
No. 2]
Adverse Possession for Intellectual Property
391
compulsory licensing to combat patent nonuse or suppression434 is
thus completely separate.
Such a distinction should sufficiently insulate our proposal from
the concern that it contravenes established Supreme Court jurisprudence that “[t]here is nothing in the patent law that requires working
of a patented invention and the general rule is that a patentee need not
use its inventions.”435 In Continental Paper Bag Co. v. Eastern Paper
Bag Co., the Court upheld a patentee's right to “accumulate patents
merely for the purpose of protecting [its] general [industry] and shutting out competitors” in order to make more money with prior patented technology.436 Numerous cases have subsequently upheld the
right not to use a patent.437
VI. CONCLUSION
Other scholars have attempted to address the conflict between
Xerox and Kodak II from various perspectives, 438 but we believe that
our proposal provides a more elegant model for addressing the
broader question of how to evaluate refusals to license not just patents
but also copyrights and trade secrets. Our proposal triangulates between the conflicting positions on the treatment of intellectual property in the shadow of the antitrust regime contained in the Kodak II,
Xerox, and Microsoft opinions, and it stakes out a position that appropriately protects competition while still respecting the rights of the
intellectual property owner.
434. See id. at 434.
435. Sobel, supra note 236, at 691.
436. 210 U.S. 405, 428 (1908).
437. See, e.g., Dawson Chem. Co. v. Rohm & Haas Co. 448 U.S. 176, 215 (1980) (discussing the “long-settled view that the essence of a patent grant is the right to exclude others
from profiting by the patented invention”); Special Equip. Co. v. Coe, 324 U.S. 370 (1945);
United Shoe Mach. Corp. v. O’Donnell Rubber Prods. Co., 84 F.2d 383, 386 (6th Cir. 1936)
(“Whatever may be the policy of patent laws elsewhere, it has long been settled that, in the
United States, exclusion of competitors is the very essence of the right conferred by a patent, and it is the privilege of any owner of property to use or not to use it without question of
motive.”); Crown Die & Tool Co. v. Nye Tool & Mach. Works, 261 U.S. 24, 35 (1923) (A
patentee may refuse “to make, use or vend his patented invention . . . .” The “essence” of
“the patent was the right to exclude others . . . .”); Motion Picture Patents Co. v. Universal
Film Mfg. Co., 243 U.S. 502, 514 (1917) (Under the patent system, a patent owner clearly
had the right to “withhold his patent[ed] machine from public use . . . .”); Lewis Blind Stitch
Mach. Co. v. Premium Mfg. Co., 163 F. 950, 954 (8th Cir. 1908) (“[D]uring the life of his
monopoly, a patentee is under no obligation to use or place upon the market . . . his invention.”).
438. See, e.g., Mark R. Patterson, When Is Property Intellectual? The Leveraging Problem, 73 S. CAL. L. REV. 1133 (2000); Carrier, supra note 297; Jonathan I. Gleklen, Per Se
Legality for Unilateral Refusals to License IP is Correct as a Matter of Law and Policy,
THE ANTITRUST SOURCE (July 2002), available at http://www.abanet.org/antitrust/
source/july02/gleklen.pdf; Jeffrey K. MacKie-Mason, Antitrust Immunity for Refusals to
Deal in (Intellectual) Property Is a Slippery Slope, THE ANTITRUST SOURCE (July 2002),
available at http://www.abanet.org/antitrust/source/july02/mackie.pdf.
392
Harvard Journal of Law & Technology
[Vol. 16
While philosophically closest to the Federal Circuit’s position in
Xerox, our proposal starts with the premise that intellectual property
owners should be rightfully entitled to claim the full scope of their
property grant, and that any activity within the scope of that grant is
permissible. We qualify that original premise, however, by introducing a concept of adverse possession for intellectual property as a restriction on certain activities that would otherwise be within the scope
of the intellectual property grant. We justify our application of adverse possession to intellectual property using many of the same arguments that are used for real property, including reliance, moral, and
Lockean justifications.439
Our proposal dictates that intellectual property owners should be
entitled ab initio to exclude other firms from developing ancillary
markets within the scope of the original intellectual property grant, via
unilateral refusals to license the intellectual property. If, however, the
other firms are not excluded and ancillary markets develop and provide a product or service that is separately demanded by consumers,
then, after a sufficient period of time, adverse possession would preclude the original intellectual property owners from excluding other
firms from the ancillary markets as they pertain to existing products.440 The intellectual property owner would, however, be entitled to
foreclose the development of new ancillary markets relating to new
products unless (1) the intellectual property owner again fails to claim
the entire scope of the intellectual property grant relative to potential
ancillary markets, or (2) the intellectual property owner has market
power and the underlying product is one that is characterized by network effects. In the latter case, we propose an application of the Rule
of Reason if the underlying product is found to be an essential facility.
One of the requirements that we imposed on ourselves in crafting
our proposal was that the resulting test be sufficiently objective so as
to be appropriately applied at summary judgment. Resolving such
cases at summary judgment is a more efficient use of judicial resources than the intent-based inquiry proposed in Kodak II, which
requires a jury to subjectively evaluate the intent of the intellectual
property owner. Given that the Sixth Circuit upheld the district court’s
summary judgment ruling in PSI Repair Services. v. Honeywell,
Inc.441 using a rationale that is quite consistent with our proposal, we
are confident that our test can be applied at summary judgment.
Rather than merely stating a theoretical case for our proposal, we
have also demonstrated how our proposal would operate in various
439. See supra Part IV.A.1.
440. Consistent with our desire to protect both competition and consumers, the prohibition against excluding other firms would include both current occupants as well as future
entrants.
441. 104 F.3d 811, 822 (6th Cir. 1997).
No. 2]
Adverse Possession for Intellectual Property
393
market scenarios. We examined, both for firms with and without market power, how our proposal would resolve the situations presented by
large fixed asset purchases, the introduction of entirely new products,
and operating systems with network effects. We have even demonstrated how our proposal could be applied in the European antitrust
enforcement context. Thus, our proposal adheres to the admonishment
in Kodak I that any test must be objectively based on market realities
rather than “[l]egal presumptions that rest on formalistic distinctions.”442
Questions to be considered later include whether a company that
would otherwise like to integrate its products should be required to
provide unbundled versions so that consumers wanting to buy one
product but not another would have the ability and incentive to do so.
In other words, again taking the Microsoft example, should Microsoft
be required to sell a version of Windows without an Internet browser
or other piece of application software? One way to analyze this question might be to use the approach the D.C. Circuit took in Microsoft
1998 when it asked whether or not the overall benefit of integrating
the products exceeded the cost to consumers of not being able to
choose the individual component parts. Or, building on the D.C. Circuit’s rejection in Microsoft 2001 of per se treatment of tying as it
pertains to operating systems, we would urge the U.S. Supreme Court
to consider abandoning the façade of per se treatment of tying arrangements and replacing the tortured analysis currently used to determine whether there are two separate products with a straight
forward application of the Rule of Reason.
442. Kodak I, 504 U.S. 451, 466–67 (1992).